Super-resolution (SR) techniques have made major advances in reconstructing high-resolution images from low-resolution inputs. The increased resolution provides visual enhancement and utility for monitoring tasks. In particular, SR has been increasingly developed for satellite-based Earth observation, with applications in urban planning, agriculture, ecology, and disaster response. However, existing SR studies and benchmarks typically use fidelity metrics such as PSNR or SSIM, whereas the true utility of super-resolved images lies in supporting downstream tasks such as land cover classification, biomass estimation, and change detection. To bridge this gap, we introduce GeoSR-Bench, a downstream task-integrated SR benchmark dataset to evaluate SR models beyond fidelity metrics. GeoSR-Bench comprises spatially co-located, temporally aligned, and quality-controlled image pairs from about 36,000 locations across diverse land covers, spanning resolutions from 500m to 0.6m. To the best of our knowledge, GeoSR-Bench is the first SR benchmark that directly connects improved image resolution from SR models with downstream Earth monitoring tasks, including land cover segmentation, infrastructure mapping, and biophysical variable estimation. Using GeoSR-Bench, we benchmark GAN, transformer, neural operator, and diffusion-based SR models on perceptual quality and downstream task performance. We conduct experiments with 270 settings, covering 2 cross-platform SR tasks, 9 SR models, 3 downstream task models, and 5 downstream tasks for each SR task. The results show that improvements in traditional SR metrics often do not correlate with gains in task performance, and the correlations can be negative, indicating that these metrics provide limited guidance for selecting superior models for downstream tasks. This reveals the need to integrate downstream tasks into SR model development and evaluation.
We introduce DORMS (Depth-Ordered Region Merging for Segmentation), a purely algorithmic approach to image segmentation from monocular depth maps, requiring no neural networks or learned parameters in the segmentation process. An off-the-shelf monocular depth estimator provides the input depth maps. Our method constructs depth-ordered merge trees through a deterministic union-find sweep, combining component area, depth persistence, and spatial contact constraints to guide region merging. A threshold search determines the segmentation granularity according to a specified target region count, while global depth detrending improves the separation of local geometric structures. Extensive experiments on COCO Panoptic, Cityscapes, and ADE20K demonstrate consistent improvements over classical segmentation baselines and competitive performance against zero-shot neural segmentation models. Notably, our method outperforms all evaluated zero-shot models on Cityscapes, while remaining competitive with several approaches on COCO and ADE20K. These results highlight the potential of purely algorithmic geometric reasoning for image segmentation without learned segmentation models.
Multi-institutional grading of Barrett's esophagus dysplasia without stain normalization using a pathology foundation model and multiple instance learning
Amirali Khosravi, Jaidip M. Jagtap, D. Chamil Codipilly, Catherine E. Hagen, Jason T. Lewis, Rickey E. Carter, Shahriar Faghani, Vani J. A. Konda 等 14 位作者
Barrett's esophagus dysplasia grading suffers substantial inter-observer variability, particularly for low-grade dysplasia, complicating clinical decision-making. We conducted a retrospective multi-institutional diagnostic accuracy study of a foundation model-based deep learning system for dysplasia grading, using 969 H\&E-stained whole slide images from four institutions. Patch-level classifiers trained on expert annotations identified diagnostically relevant regions, followed by attribute-aware multiple instance learning for slide-level classification as non-dysplastic Barrett's esophagus (NDBE), low-grade dysplasia, or high-grade dysplasia. On a held-out test set (n = 130), macro-averaged AUC was 0.96 (95\% CI 0.93--0.98). For binary dysplasia detection, sensitivity was 98\% (80/82) and specificity 81\% (39/48), with a negative likelihood ratio of 0.030 and quadratic-weighted \kap{} of 0.83 against expert consensus. No high-grade dysplasia case was misclassified as NDBE. Performance was consistent across institutions without stain normalization, indicating that stain-robust foundation-model representations may remove the site-specific preprocessing that has constrained prior systems to institution-specific deployment.
When multimodal decision models determine actions, moderation outcomes, or quality judgments, their reliability depends on preserving semantic decisions across presentation changes, ignoring irrelevant additions, and distinguishing information consistent with task evidence from information that conflicts with it. Existing evaluations study decision-interface biases and multimodal model behavior in separate settings, leaving no common protocol to compare multimodal decision models with their backbones on practical tasks. We introduce \textbf{MM-SystemOneBias}, which evaluates 13 models on 5,323 original instances from nine public datasets and 11 task scenarios using six paired interventions: option order, option label, external judgment, image--text description, irrelevant information, and rating scale. Across seven matched decision-model--backbone pairs, decision models are more stable under option-label, irrelevant-information, and rating-scale changes in all seven pairs; six pairs also show lower option-order flip rates. They incur smaller accuracy losses under incorrect external judgments in all seven pairs and conflicting descriptions in six. In six pairs, they gain less accuracy from correct external judgments and show less favorable accuracy changes under consistent descriptions. Among pairs with unchanged semantic decisions, mean absolute changes in answer probabilities range from 4.56 to 11.55 percentage points, revealing input effects that accuracy alone misses.
This preprint presents a systematic, research-oriented guide to modern natural language processing, covering the full NLP pipeline from tokenisation, text normalisation, vector representations, and classical machine learning to transformer architectures, large language models, retrieval-augmented generation, preference-based alignment, and multimodal systems. The material comprises eighteen practical works combining theoretical foundations, reproducible implementation examples, detailed experimental procedures, evaluation metrics, and research-oriented assignments. Each work includes three structured core tasks and optional advanced investigations. Common requirements for reporting, data provenance, experimental evaluation, and reproducibility are established across the guide. Particular attention is devoted to low-resource and morphologically rich languages, including Tajik, Tatar, and Bashkir. The guide addresses subword tokenisation, cross-lingual transfer, machine translation, language model adaptation, and the evaluation of NLP systems under limited data and computational resources. Advanced topics include prompt engineering, reasoning strategies, multi-agent collaboration, model compression, efficient inference, AI safety, and LLMOps. A capstone project integrates methods from different chapters into the development and evaluation of domain-specific NLP systems. The guide is intended for senior undergraduate students, graduate students, researchers, and developers interested in implementing and critically evaluating NLP methods.
Wildfires represent a critical ecological and socio-economic challenge in Canada, where vast boreal forests and a warming climate have amplified the frequency and severity of extreme fire seasons. Predicting next-day wildfire spread can support fire management, but requires capturing complex interactions among weather, fuels, and terrain. Deep learning has emerged as a promising approach to learning these relationships from heterogeneous environmental data, yet its development and evaluation depend on representative datasets. Existing deep learning research on wildfire spread relies predominantly on US-focused datasets, leaving Canadian fire environments underrepresented. To address this gap, we present DailyWildfireCAN, a multi-modal, multi-temporal dataset and benchmark for predicting next-day wildfire spread across Canada. The dataset tracks continuous fire progression over multi-day events at a high spatial and temporal resolution of 90-m and 24-hour, respectively. This resolution preserves fine-scale information relevant to fire-edge geometry, local fuel variations, and terrain effects, supporting research toward improved daily wildfire spread forecasting. DailyWildfireCAN covers 2,440 wildfire events across eight fire seasons 2017-2024 and integrates 33 features across diverse modalities, including topography, meteorology, land cover, vegetation and forest structure, human influence indices, and satellite imagery. We establish a comprehensive benchmark comparing mono-temporal and multi-temporal prediction models, accompanied by a multi-dimensional evaluation framework assessing robustness across wildfire characteristics, geographic regions, and temporal conditions, spatial prediction quality, and sensitivity to reference label variations.
Linear probes are the cheapest LLM lie detectors but often fail to transfer across kinds of lie. Label-free directions from honest and deceptive role-play are untested at 70B scale and on lies about knowledge, where a lie and an honest denial can be identical. We widen a published persona bank with roles that deny knowledge they have, plus honest counterparts. At each model's middle layer the resulting persona direction ranks lies across Liars' Bench better than single-dataset probes on three of its four models; it fails on Qwen, even when Qwen stays in role far more often. On Llama-3.3-70B-Instruct it beats directions from a random persona bank, but its gain over the published direction is within the range of random widenings. On lies about knowledge, persona directions mainly track a reply's wording: where lies and honest denials share their words, our three main contrasts stay inside the random-direction band on Llama and Qwen and reach only its edge on Gemma, though on Mistral one clears it. A knowledge direction contrasts questions the model can answer with ones nobody can, each followed by the same scripted denial. It separates lies from identical honest denials on two models and narrowly on a third, but not on the fourth; its other results partly depend on the benchmark's wording. On the Gender-Secret fine-tune it clears the band by a few hundredths on Llama and fails on Mistral. The sum of the two directions ranks lies better than a supervised probe at the same depth on all four models (on one, both near chance) but is level with the persona direction on Llama's average and behind that probe on held-out lies. More varied role-play questions do not help the persona direction on Liars' Bench.
LipSpell: A Language-Independent Phonetic Scoring Framework for Spelling Correction Combining Distinctive-Feature Edit Distance, Keyboard Geometry, and Corpus Frequency
Automatic spelling correction must handle both keyboard typographical slips and cognitive or phonetic misspellings where writers approximate a word's pronunciation. Conventional spell checkers rely either on orthographic edit distance and keyboard proximity--which struggle when phonetic errors diverge by multiple characters--or on English-centric phonetic hashes such as Soundex and Metaphone. We present LipSpell, a unified, language-independent phonetic scoring engine paired with swappable language-specific G2P frontends, keyboard maps, and lexicons integrating: (1) a precomputed |P|x|P| phone-distance matrix from universal IPA distinctive features, (2) physical keyboard geometry, and (3) corpus unigram frequency priors, accelerated by a two-stage symmetric-delete and seven-class phonetic-code index. Across six English benchmarks (including 1,000 L2 errors from TOEFL-Spell), LipSpell outperforms Norvig and Hunspell on every dataset and beats Aspell on Norvig1 (78.52% vs. 75.56%), Holbrook (32.00% vs. 26.67%), and TOEFL-Spell (77.30% vs. 72.80%). Evaluated across six languages spanning five branches (Germanic, Romance, Slavic, Semitic, Turkic), three families (Indo-European, Afro-Asiatic, Turkic), and three scripts, LipSpell achieves strong fixed-weight transfer on Brazilian Portuguese (73.00% Top-1 on n=500 Wikipedia without NFD tolerance and 75.80% with NFD vs. 66.40% Norvig, 64.20% Hunspell, 64.00% Aspell) and Russian Cyrillic (85.67% vs. 78.50% Norvig). On 50/50 clean non-word splits with lexicon cleaning and weak-character indels, tuning yields strong gains over Norvig in Portuguese (86.97% vs. 76.75%), Turkish (81.67% vs. 79.05%), Hebrew Abjad (74.59% vs. 62.30%), and French (74.53% vs. 67.92%), while fixed weights remain optimal for English (80.37%) and Russian (85.67%).
Single-molecule force spectroscopy (SMFS) probes biomolecular mechanics at the single-molecule level, but high-throughput acquisition of force-extension curves creates a data processing bottleneck. Identifying rare specific unbinding events among thousands of noise-dominated traces still relies largely on manual visual inspection. Here, we present a deep learning workflow for the preliminary screening of severely imbalanced SMFS datasets. Using 2D image representations of force-extension curves without system-specific physical priors, we fine-tuned a modified ResNet18 architecture trained with focal loss and evaluated it on the mechanical unfolding pathways of the Ruminococcus champanellensis XMod-Doc/Coh complex. On a held-out test set with a 1.34% target prevalence (13 positive P_2 events among 970 curves), the model achieved 92.31% recall (12 of 13 events) and 91.95% specificity (880 of 957 negative traces), automatically discarding 90.82% of the dataset and reducing manual review from 970 to 89 flagged curves. We also integrated Gradient-weighted Class Activation Mapping (Grad-CAM) to provide a qualitative visual reference of the curve regions driving each prediction. The complete Python pipeline is released as open-source software with an executable Google Colab notebook, enabling reproducible SMFS screening without dedicated local GPU hardware.
We present a longitudinal measurement study on the adoption of detectable, second-generation anonymisation protocols in the Bitcoin network, including CoinJoin, CoinSwap, CoinShuffle and Stealth Addresses. By implementing and refining a suite of heuristic filters, we identify over 5 million candidate CoinJoin and 23.3 million candidate CoinSwap transactions. Notably, the use of CoinShuffle was unexpectedly found to be closely aligned with the operation of coordinator-based CoinJoin services such as Wasabi and Whirlpool. Our analysis reveals consistently low adoption rates, with these protocols typically constituting less than 1% of network transactions, and a sharp decline in detectable usage following key regulatory events. Furthermore, we find no evidence of standardised Stealth Address adoption, indicating a failure to converge on a common privacy standard. This study provides a comprehensive picture of a niche ecosystem whose on-chain visibility has been largely suppressed, strongly suggesting the migration of privacy-seeking users to less transparent and less detectable methods.
Intersectional bias arises when a model treats people who hold several marginalised identities at once differently from what their separate identities would predict. Auditing such intersectional subgroups remains a challenge: their number grows quadratically, and there is no agreed baseline for what independent attributes would predict. Without such a baseline, many existing audits flag a subgroup when it exceeds each attribute alone, a condition that attributes acting independently already satisfy. We test this standard criterion on 150 synthetic datasets built with no interaction. With the shuffle null, it flags 66.1 false subgroups per dataset on average, and at least one in every dataset. When we plant interactions of known size into these datasets, a repaired null lets the criterion recover only 32% of them. We introduce ADDIS (Additive Intersectional Screen), which fits an additive model over all attributes, scores each subgroup by its Additive Residual Distance (ARD) from that model's prediction, and tests it against a Freedman-Lane null. On our synthetic benchmark, ADDIS flags 0.09 false subgroups per dataset and recovers 84% of the same planted interactions. We then build IndiPersona, 10,000 synthetic Indian personas with six identity attributes, and audit 13 open language models. The standard criterion flags 7.1% of the 738 tested subgroups on average; ADDIS flags 3.0%. Gender identity crossed with sexual orientation is the largest pairing in every model, and counterfactual personas confirm it and additionally reveal a caste x class interaction in all but one model that the observational test lacked the data to detect. Much of what audits report as intersectional bias can thus arise from independent single-attribute effects alone, yet a small core of interactions survives.
Multivariate time series forecasting (MTSF) is critical across many real-world domains. Existing deep learning approaches fall into two paradigms with distinct limitations: channel-independent (CI) methods ignore cross-variable dependencies and model only temporal dynamics, while channel-dependent (CD) methods consider both but typically rely on architectural compromises to mitigate overfitting and computational overhead. We therefore propose Chameleon, a CD state space model (SSM) that enables data-dependent, fine-grained interactions across variables while scaling linearly with the number of variables. By connecting selective SSMs with the Kalman filter, we adapt the latter's measurement-update formulation for cross-variable modeling while preserving the SSM backbone for robust temporal modeling. We further identify favorable inductive biases of GatedDeltaNet for time series, adapt it as our backbone, and improve generalization through additional techniques, including a previously unexplored stochastic perturbation of reversible instance normalization. On strongly dependent ODE and PEMS datasets, Chameleon achieves the best MSE and MAE across all settings, while its CI ablation and prior CD methods incur 61-178% higher MSE on average. Across 28 standard benchmark settings, Chameleon also achieves better MSE and MAE than each baseline in at least 27 and 22 cases, respectively. Training-time, FLOP, and peak-memory analyses on Traffic and ETT further demonstrate competitive efficiency and favorable scalability across different variable counts.
Real applications provide the training setting closest to phone-agent deployment, but are difficult to reset, scale safely, and verify programmatically. Static screenshots and interaction trajectories preserve realistic evidence but cannot generate new experience. We introduce PhoneWorld, a trace-grounded framework that converts such evidence into runnable, resettable, and verifiable Android environments. PhoneWorld induces a usage-weighted interaction skeleton from observed pages, transitions, and state-changing operations; translates it into a behavior-grounded app specification; realizes the specification through an autonomous build--inspect--repair loop; and synthesizes executable tasks with programmatic verifiers. The resulting suite spans 34 consumer-facing apps across 16 domains and supports an audited online benchmark, verified trajectory generation, and online RL through common reset and verification interfaces. Evaluations with diverse general and open-source GUI agents show that PhoneWorld supports reliable end-to-end online interaction and exposes capabilities complementary to AndroidWorld. Controlled SFT experiments further show that PhoneWorld trajectories complement AndroidWorld supervision, transfer across online and offline benchmarks, and become more effective as data volume and app coverage increase. Under a matched RL budget, combining PhoneWorld mock-app rollouts with real-app rollouts improves performance over real-app RL alone on both real-phone tasks and AndroidWorld. Together, these results demonstrate that trace-grounded executable abstraction can bridge realistic mobile behavior and scalable agent learning, turning limited real-app evidence into a growing supply of controllable and verifiable environments for training and evaluation.
Real-world agents operate over long and evolving horizons, where information is repeatedly updated and may interfere across memories, requiring accurate recall and aggregated reasoning over multiple pieces of information. However, existing benchmarks focus on static, independent recall and fail to capture these dynamic interactions between evolving memories. In this paper, we study how current memory-augmented agents perform in realistic, interference-heavy, long-horizon settings across diverse domains and question types. We introduce MINTEval (Long-Horizon Memory under INTerference Evaluation), a benchmark featuring (1) long, highly interconnected contexts with frequently updated information that induces substantial interference, (2) diverse domains (state tracking, multi-turn dialogue, Wikipedia revisions, and GitHub commits), enabling evaluation of domain generalization, and (3) diverse question types that assess robustness to interference, including (i) single-target recall tasks requiring retrieval of a specific target from long contexts, and (ii) multi-target aggregation tasks requiring reasoning over multiple relevant pieces of information. Overall, MINTEval has 15.6k question-answering pairs over long-horizon contexts averaging 138.8k tokens and extending up to 1.8M tokens per instance. We evaluate 7 representative systems, including vanilla long-context LLMs, RAG, and memory-augmented agent frameworks. Across all systems, we observe consistently low performance (avg. 27.9% accuracy), especially on questions requiring aggregated reasoning over multiple pieces of evidence. Our analysis shows that performance is primarily limited by retrieval and memory construction. Furthermore, current memory systems struggle to recall and reason over earlier facts that are revised or interfered with by subsequent context, with accuracy degrading as the number of intervening updates increases.
Complete test suites for automata are (finite) sets of words that enable testing an implementation against a specification: when a (black-box) automaton passes the test suite, completeness guarantees that it is equivalent to the specification on all (infinitely many) words. A well-known method for constructing complete test suites for deterministic automata is the W-method, used in conformance testing and automata learning. Based on a recent categorical generalisation of the W-method, we propose a framework for~the construction and analysis of test suites for a certain class of effectful automata defined by commutative monads. By working in the Eilenberg-Moore category of the associated monad at hand, we obtain a generic notion of effectful W-method, together with a completeness theorem. We show how to apply the framework and derive complete test suites for non-deterministic and weighted automata.
Robust Pedestrian Detection with Uncertain Modality
Qian Bie, Xiao Wang, Bin Yang, Zhixi Yu, Jun Chen, Xin Xu
AI 速读 基于摘要
作者构建RGB-NIR-TIR三元组数据集并提出AUNet,以在模态组合不确定时实现鲁棒行人检测。
cs.CV
查看原文摘要
Existing cross-modal pedestrian detection (CMPD) employs complementary information from RGB and thermal-infrared (TIR) modalities to detect pedestrians in 24h-surveillance systems.RGB captures rich pedestrian details under daylight, while TIR excels at night. However, TIR focuses primarily on the person's silhouette, neglecting critical texture details essential for detection. While the near-infrared (NIR) captures texture under low-light conditions, which effectively alleviates performance issues of RGB and detail loss in TIR, thereby reducing missed detections. To this end, we construct a new Triplet RGB-NIR-TIR (TRNT) dataset, comprising 8,281 pixel-aligned image triplets, establishing a comprehensive foundation for algorithmic research. However, due to the variable nature of real-world scenarios, imaging devices may not always capture all three modalities simultaneously. This results in input data with unpredictable combinations of modal types, which challenge existing CMPD methods that fail to extract robust pedestrian information under arbitrary input combinations, leading to significant performance degradation. To address these challenges, we propose the Adaptive Uncertainty-aware Network (AUNet) for accurately discriminating modal availability and fully utilizing the available information under uncertain inputs. Specifically, we introduce Unified Modality Validation Refinement (UMVR), which includes an uncertainty-aware router to validate modal availability and a semantic refinement to ensure the reliability of information within the modality. Furthermore, we design a Modality-Aware Interaction (MAI) module to adaptively activate or deactivate its internal interaction mechanisms per UMVR output, enabling effective complementary information fusion from available modalities.
End-to-end driving policies can achieve strong task performance while offering limited control over their driving style. Driving style describes how a vehicle accelerates, brakes, and maintains distance. Policies trained without explicit preference conditioning lack a direct mechanism for adjusting these behaviors at inference. We propose YOUDrive, a vision--language--action planner for continuous and composable driving-style control. Each teacher's behavior is distilled into a task vector, a learned parameter update that induces a style direction in the planner's representation. These directions can be scaled and combined at inference without retraining the backbone. Conditioned on these directions, a flow-matching decoder models a multimodal trajectory distribution and generates kinematically feasible motion alternatives. We introduce the Style Alignment Score (SAS) to describe human-referenced kinematic behavior after safety and task gating. On NAVSIM navtest, a single coefficient monotonically controls assertiveness across six teacher-derived personas, while PDMS remains above 0.90 across the tested individual and blended configurations, reaching 0.954. Further analyses reveal systematic policy-specific kinematics after controlling for scene context, whereas the evaluated discrete style labels provide limited additional explanatory power. Code, checkpoints, and the metric will be released.
Large Web platforms operate large fleets of ranking models, each selecting features from a shared pool of tens of thousands. We present an end-to-end agentic system that optimizes features across the fleet under resource constraints shared across models and those set per model, such as feature pipeline cost and serving throughput (QPS), respectively. In this system, LLM agents propose hypotheses, and deterministic tools execute experiments and verify them. For each model, an LLM generates and shortlists cross features from a combinatorial candidate space, and all features are then ranked by several importance methods. Feature blending mixes these rankings into a single selection, using Pareto fronts across tasks for multi-task models. An LLM Researcher agent then tunes the feature selection in 10--25 asynchronous trials under resource caps, and a trimming step keeps the pipeline cost within budget. Across production models, the system improves normalized entropy by 0.31% on average within each model's budget, and the trimming step cuts pipeline cost by 2.3% with no regression beyond measurement noise. The optimized models were launched, and in online A/B tests they improved a primary online metric by 0.20%.
Inverse RNA Folding consists of finding a nucleotide sequence that folds into a predefined secondary structure. It is useful for synthetic biology, medicine, and nanotechnology. We propose Montparnasse, a Monte Carlo search framework based on Generalized Nested Rollout Policy Adaptation, augmented with a structural prior derived from the energy model, slow and long adaptation at level 1, a lexicographic multicriteria evaluation, a local repair of the pairs the predicted folding gets wrong, and a fleet of parallel workers coupled through a shared beam whose diversity is maintained on the faults the candidates make rather than on the candidates themselves. On the Eterna100 V1 benchmark, with both solvers folding through the same ViennaRNA 1.8.5 library under the Turner 1999 model, with 100 parallel workers and a budget of 30 minutes per puzzle, Montparnasse solves a median of 96 of the 100 puzzles against a median of 90 for DesiRNA, the previous state of the art, each figure taken over five passes. It is faster on 88 of the 90 puzzles that both solve, with a median solving time of 0.9 s against 5.6 s, and it solves 55 puzzles within the first second. The one puzzle no pass reaches is solved under the same model with a longer budget, so every Eterna100 V1 target is solved by Montparnasse.
Precise grounding matters. It specifies which object is the target and where that object is, even in clutter and for tiny objects, and it has to be fast enough for closed-loop control. Yet vision-language-action (VLA) and world-action models (WAMs) take perception from general-purpose vision-language and video-generation backbones, which still fail in these settings. We introduce GroundingPI, a 4B grounding foundation model that generates points and boxes as quantized coordinates in a shared vocabulary. Training combines multimodal and spatial pretraining, supervised fine-tuning, and reinforcement learning with GRPO, using supervision from public datasets and dedicated data engines. Against 44 baselines across 34 grounding benchmarks spanning 11 perceptual capabilities, GroundingPI establishes a new state of the art, averaging 73.68%, above the larger GPT-6 Astra (71.54%). As a downstream visual backbone, GroundingPI improves performance on robotic manipulation and autonomous driving. Across RoboTwin 2.0 and RoboCasa-GR1, it outperforms all evaluated backbones in all four out-of-distribution settings, with relative gains of up to 24.8% over the strongest baseline. On RoboCasa-GR1, GroundingPI trained with 50% of the demonstrations outperforms those baselines trained with 75%. Its mean open-loop L2 error on nuScenes is 0.296 m. We systematically analyze GroundingPI's pretraining in scale and data composition. Grounding quality and downstream manipulation improve with pretraining scale, with generalization tracking perceptual gains most closely. Analyzing the data recipe across these 11 perceptual capabilities shows dense grounding's substantial benefits for both, and OCR's potential as a catalyst for perceptual learning. These results support grounding as a perceptual foundation, and dedicated perceptual pretraining as a promising direction for foundation models of physical intelligence.
A benchmarking campaign provides the measurements behind an HPC performance study: code variants are built, swept over core or node counts, and measured repeatedly. Moving such a campaign to another supercomputer or rerunning it months later requires work that is not part of the study itself. Job scripts are adapted to each supercomputer, and what exactly ran is often reconstructed afterwards. We present Thinflow, a small Bash engine for organizing HPC benchmarking campaigns as rerunnable directory trees of tasks and runs. A task is a directory with a short shell script next to the code. For each sweep, one command builds the code, runs the benchmarks, and plots the results. Thinflow derives the execution order from declared dependencies and writes it to a plain-text manifest. Per-supercomputer profiles submit the runs to a workload manager such as Slurm. Each run writes a provenance record of its environment, hardware, and parameters. The engine ships inside the campaign directory, so a campaign stays rerunnable without installing a separate tool and, once completed, may serve as a paper artifact. In a case study, we ran one campaign with identical task files on three Slurm partitions of two supercomputers, on up to 16 nodes. This strong-scaling study of matrix multiplication in C/OpenMP, C/MPI, and Julia/MPI comprises 37 task files with 217 non-comment lines. Its 455 benchmark runs recorded what differed per partition, chiefly profiles and command-line parameters. This suggests that such a campaign needs little structure beyond plain shell files.
Agent applications are increasingly deployed, yet agent-specific attacks pose real threats to them. Upon a security incident, organizations backtrack to identify the CEP (compromise entry point), remediate the damage, and prevent its recurrence. The main artifacts available for backtracking are trajectory logs, which record agent functionalities, and source code; app(lication)-specific function call logs are often fairly limited. However, trajectory logs are at the agent functionality abstraction layer, while the CEP candidates are function calls at the app code execution layer. Due to this semantic gap, the analyst needs to manually inspect the relevant code for each CEP candidate to infer its corresponding agent functionality. In this work, we present Origin, a backtracking tool for the analyst to submit a small number of queries to identify CEPs that existing tools cannot provide. Our approach poses two challenges: (1) define the intermediate abstraction that captures the mapping relationships between the two semantic layers; (2) determine the region-to-variable mapping for constructing the cross-layer graph defined in the intermediate abstraction. Each region is a specific group of agent functionalities. Origin addresses the challenges by (1) the cross-layer dependency graph (CLDG) that connects region-inferred functions at the semantic level to the variables in the low-level taint propagation subgraph, plus the associated query language, and (2) the region inference algorithm. Case studies on five popular agent applications across eight attack incidents show that an analyst using Origin correctly identifies all the CEPs.
Reliable extraction of calcium-dependent fluorescence traces in calcium imaging sequences requires identifying pixels whose fluorescence intensity reflects the same underlying cellular activity. We present a deterministic temporal embedding framework that maps pixel-wise fluorescence traces to a low-dimensional representation in which temporally coherent sources form spatially coherent regions. Starting from a fuzzy-neighborhood graph and spectral initialization, we optimize a class-balanced cross-entropy objective using exact sparse attraction and a deterministic particle-mesh decomposition of the dense repulsive interaction with sparse edge compensation, avoiding stochastic negative sampling. The resulting latent representation is spatially decoded into reliable source supports for trace extraction without imposing an explicit cellular shape prior. Experiments on synthetic and in vivo two-photon calcium imaging data demonstrate robust recovery of irregular and spatially touching sources together with efficient deterministic embedding optimization.
In our experiments on long-horizon customer-behavior forecasting, a longer observation window outweighed the observed effects of more data, more capacity and newer architectures. The window is the weeks of history a model sees before the week it forecasts from. Going from an 8- to a 16-week window adds a raw +0.070 to 0.107 log-space $R^2$ in every family at 1M. Matched test cohorts, which score the same customer-weeks at every window, show this raw comparison understates the gain in every family they cover. The direction holds on five public datasets, four spanning over a year (ridge and our transformer on all five). On our internal data, longer windows admit up to 10% more and colder training customers, which matching does not fix. Our design does not separate window length from calendar coverage, because every test anchor (the week a forecast is made from) falls in one eight-week band. On training-period anchors, the weekly ridge probe's gain persists outside the holiday season, keeping about 72% of it. We therefore report a robust positive association with longer windows and do not identify the magnitude of a calendar-independent history effect. Of four internally registered predictions, the only fully blind one failed. An activity-mixture model predicted +0.014 from 24 to 32 weeks; the weekly-aggregate probe returned -0.0004. For the sequence families, the other three levers are about an order of magnitude smaller, bounded and unresolved. Their observed effects are at most 0.006 between the nominal 1M and 10M tiers and 0.0052 among five sequence families at 16 weeks with clocks as built, both single-seed on the full test split, and 0.006 for capacity on three-seed means. These are observed intervals, not equivalence bounds. Our claim is the order-of-magnitude contrast between the window and those bounds. A team building on this class of data should start by lengthening the window.
Most world models for future prediction prioritize aestetic quality of generated videos, but offer only weak physical grounding and limited geometric consistency over time, due to the lack of explicit 4D modeling. In this paper, we introduce plan4d, a possible next milestone towards building a generative plannable 4D world model that predicts future RGB videos together with dense 4D geometry, including depth, cameras, point clouds, and scene flow from an observed scene context and text-specified actions. Since training data with paired actions and dense 4D annotations is inherently scarce, plan4d leverages the complementary priors of foundational video generators and 3D reconstructors. Specifically, we propose a model-stitching architecture that couples Depth Anything 3 with LongCat-Video and augments it with an auxiliary DPT head for scene-flow prediction. The resulting model allows us to plan and generate complex, long-horizon, and fine-grained actions in a dynamic 3D scenes, while substantially improving both visual fidelity and geometric accuracy of the generated future videos. Extensive experiments on both synthetic and real-world datasets demonstrate that plan4d achieves state-of-the-art performance in generative 4D modeling of environmental reactions. Project Page: https://fogradio.github.io/plan4d-project
Improving microprocessor performance while meeting ASIC implementation constraints like area and frequency is a complex and labor-intensive task that requires expertise across computer architecture, hardware design, and performance analysis and profiling. While large teams in industry can afford the engineering resources to target this problem across performance modeling, RTL design, and ASIC implementations, small teams of researchers and engineers cannot take on this huge engineering burden while keeping up with the pace of advancement in processor microarchitecture. Recent advancements in large language model (LLM) agents present a promising substrate to accelerate this traditionally human-intensive process. In this work, we build and evaluate MiracleGro, an agentic AI-based system that, without human intervention, analyzes cycle-level RTL behavior and synthesis results, identifies performance bottlenecks, and translates architectural ideas into RTL implementation in a leading open-source out-of-order superscalar application core, the Berkeley Out-of-Order Machine (BOOM). Taking into account cycle-exact performance, area, and critical path, MiracleGro automatically produces improved BOOM designs that drastically increase the human-optimized BOOM's performance on substantial workload suites such as SPEC CPU2017 with full reference inputs. MiracleGro achieves an 85% IPC speedup and a frequency improvement of 27% in SkyWater 130nm, resulting in a 2.34x wall-clock time speedup on the full SPEC CPU2017 intspeed benchmarks (30 trillion instructions). This result demonstrates the enormous opportunity to harness frontier models to improve state-of-the-art RTL-driven platforms for implementation-grounded microarchitecture research across hardware designs.
Stochastic gradient descent (SGD) is at the center of modern machine learning and is known to be statistically minimax-optimal in many regression settings, e.g., for convex Generalized Linear Models (GLMs). Yet, it is unknown whether its simplest differentially private Noisy SGD counterpart fulfills similar favorable guarantees even on strongly convex losses. We resolve this question affirmatively. We upper bound the parameter estimation error for a single pass of Noisy SGD with a fixed stepsize schedule and constant batch sizes on smooth and strongly convex Lipschitz losses over a compact convex domain under Renyi DP. Specialized to strongly convex GLMs, our error bounds match minimax lower bounds we derive based on the Score Attack framework. In particular, this provides the first minimax-optimal rate for Noisy SGD in private logistic regression, avoiding poly-logarithmic terms present in existing results. Our analysis is versatile as it relies on standard convex analysis and establishes the same rates in expectation and with high probability.
Retrieving government schemes requires matching citizens' information needs to formal administrative descriptions, often across languages and terminology. We present a bilingual search system for Maharashtra government schemes and public services that combines BM25 retrieval with sentence-embedding similarity through equal-weight score fusion. The system indexes aligned English-Marathi scheme descriptions and supports multilingual, Marathi-focused, and translation-mediated semantic configurations. Evaluation on manually curated bilingual queries compares BM25, dense retrieval, and hybrid retrieval using IndicSBERT and MahaSBERT. BM25 provides a strong baseline with a mean reciprocal rank (MRR) of 0.864. Hybrid retrieval increases MRR to 0.872 with IndicSBERT and 0.879 with MahaSBERT, whereas the corresponding dense-only systems achieve 0.663 and 0.655. MahaSBERT-based fusion also increases precision from 0.781 to 0.808 while matching BM25 recall at 0.946. These results support the complementary use of lexical and semantic signals in this collection, while showing that improvements over a strong lexical baseline are modest and metric-dependent. The study provides an application-focused analysis of hybrid retrieval for bilingual public-service information access.
Native 3D generation has largely been built around latent diffusion, with recent advances concentrating on stronger variational autoencoders (VAE) and increasingly elaborate latent representations, from vecset and triplanes to structured sparse voxels. Despite their architectural differences, these methods all share the same premise: geometry is first compressed into a latent space and only then modeled by the generator. We take a different route and propose Direct3D-V3, a latent-free diffusion transformer (DiT) framework that discards the VAE altogether and operates in the native voxel domain, directly modeling raw SDF voxel grids during both training and sampling. To make this direct voxel formulation geometrically reliable, we introduce an implicit manifold regularization over neighboring voxels to suppress non-watertight surfaces, and convert predicted SDF grids into meshes via Marching Cubes so that rendered normal losses can supervise the model directly from its predicted SDF. To improve alignment between the conditioning image and generated geometry, Direct3D-V3 also predicts the camera parameters of the conditioning image as an auxiliary task, encouraging the DiT to learn the mapping from image observations to 3D shape. This latent-free design removes the reconstruction bottleneck, brings training under an objective in the native voxel domain, and yields meshes with substantially richer local detail. The predicted camera and native SDF voxel output further make Direct3D-V3 extensible to generative reconstruction, where diffusion is interleaved with normal-guided refinement to improve alignment with the observed view while the generative prior completes unseen regions. Project page: https://dreamtechai.github.io/Direct3D-V3.
Task success alone does not specify which behaviors post-training should reinforce. Even successful teacher trajectories can contain unsupported assumptions, unproductive repetition, or superficial verification. We propose Skill-FT (Skill-Guided Fine-Tuning), which uses natural-language skills to suppress undesirable habits and elicit underrepresented effective behaviors during teacher rollouts, then fine-tunes a target model on the resulting trajectories. We introduce AutoMaster, an autonomous post-training system in which an LLM agent develops these interventions through trajectory analysis, skill design, data selection, and feedback from rollouts and target-model evaluation. Across Terminal, Search, and software engineering tasks with 4B and 9B target models, AutoMaster improves mean performance over Vanilla SFT on successful unguided teacher trajectories by 1.42-8.99 percentage points, with the teacher, task pool, training set size, and SFT recipe held fixed. Together with the ablations and trajectory analyses, these results suggest that behavioral shaping can make a fixed model a more effective teacher. We further find that retaining skill instructions during SFT yields higher mean performance than removing them across all six settings, despite their absence at evaluation and identical supervised outputs. These findings suggest that preserving skill context as part of a training demonstration can support better behavioral internalization.