Interpretability, AI auditing and assessments of machine consciousness increasingly rest on mechanism-level quantities: functionals of a system's causal transition law that describe how it computes, not what it outputs. Experiments see that law only through finite, partial and noisy interventions. We ask when such a quantity is identifiable from an experiment's access, continuous in the mechanism and certifiable from finite data, and answer exactly for a deliberately demanding case - the integrated information $\varphi$ of integrated information theory - in stochastic networks of four and five units with known ground truth. Behavior can be silent: a recurrent and a feed-forward network with identical input-output laws at their response delay carry 0.28 and 0 bits, yet internal recordings separate them within 4-53 trials. Partial access can be just as blind: in our examples, units left out of a one-step recording impose precision floors of 0.30-1.02 bits that no sample size lowers, and in all-but-self parity networks, recordings of any length see several hidden units only through the product of their fidelities - a symmetry that one intervention broke in exact checks. The measure has sharp edges: because $\varphi$ selects a partition, and the cause and effect it specifies, by one criterion and reports another, infinitesimal changes of mechanism that cross an edge change it by 25-100% at 99% of the edges we found in random networks, and about 1% of surrogate models that fit interventional data almost perfectly misreport it by more than 20%. One-sided claims survive: the same data certify a lower bound on both sides of an edge, and we prove $\varphi>1.004$ bits throughout a 64-dimensional region of mechanisms and certify it from about $10^7$ controlled transitions with false-certificate probability at most $α$. Measurability is thus a design question with computable answers.
Aamir Mahmood, Fateme Salehi, Nurul Huda Mahmood, Song Han
Predictive interference management supports ultra-reliable low-latency communication (URLLC) by allocating finite-blocklength resources against a high conditional quantile of next-slot interference. Building on an extreme-value-theory/kernel-density-estimation (EVT/KDE) predictor with quantile-defined interference states, this paper makes the predictor resilient to unannounced shifts in the interference distribution. Its cumulative refitting on all past observations does not specify how the state partition and retained samples should evolve after a regime change. We show that drift first makes the quantile partition inefficient: cumulative refitting preserves tail support and coverage but cannot recover conditional resolution, and pays for it in channel use, whereas discarding history starves the tail estimator. We therefore propose event-triggered quantile-space adaptation. A nonparametric cumulative sum (CUSUM) on the predictor's level stream detects change. At each confirmed event, the quantile partition is re-derived from post-change observations, retained history is re-binned under it, and each state is refit from a bounded per-state memory, capped to balance stale-regime bias against tail-estimation variance, with a closed-form L-moment tail estimator. At a nominal coverage target $η\!=\!0.99$, relative to matched cumulative refitting with fixed states, the proposed method reduces channel use by $5.7$--$13.2\%$ and quantile pinball loss by $25$--$50\%$ across abrupt, incremental, compound, downward, and reversal drift, with coverage changing by $-0.5$ to $+1.0$ percentage points. At matched achieved coverage, it saves $4$--$18\%$ of channel uses, while resource use in load-only and stationary cases changes by less than $0.5\%$. With $K$ samples per state and $L$ states, predictor memory is bounded at $O(KL)$ rather than growing as $O(t)$.
Mayank Anand, Subham Raj, Aman Vaibhav Jha, Sriparna Saha
Long-horizon LLM agents require memory systems that preserve useful information across sessions and retrieve it when needed. Existing systems treat these as separate problems: memories are written prospectively with static importance scores that are never revised after use. This creates two coupled failure modes: consolidation can silently destroy evidence that future queries require, while retrieval continues to favor stale memories. We propose MEMOIR (Memory Evidence via Merge-certified, Online Importance Revision), a training-free memory lifecycle that unifies judge-gated consolidation with retrospective attribution. Before compressing memory, screening gates test whether answer-relevant evidence is preserved. After each query, attribution estimates which retrieved memories supported the answer and updates their retrieval priority. A UCB-style exploration bonus additionally encourages retrieval of under-exposed memories, reducing the risk that frequently retrieved memories crowd out useful but less-seen evidence. On 1,905 deduplicated LoCoMo QA pairs, the full retrieval stack improves evidence recall@5 by 11.92 percentage points over bare cosine, with confidence-gated feedback adding 1.50 points overall without reference answers or reranker training; a deployment-unavailable oracle adds 1.84 points overall and 8.05 points on multi-hop. Judge-gated merge changes the recall of text-based evidence from 58.82% to 59.08%, reduces the active store by 1.42%, and incurs 0 observed losses in 100 independently audited merges, with a one-sided 95% upper loss-rate bound of 2.95%. Together, these results outline a practical deployment path: no additional LLM call on the online write path, with merge screening confined to periodic offline consolidation. The lifecycle reduces to one principle: attribute usefulness after use, and screen for preservation before compression.
3D Gaussian Splatting (3DGS) has become a vital tool in novel view synthesis. 3D Gaussian functions with attributes like color, opacity, and scale are stacked together to explicitly model a radiance field. However, discrete 3D Gaussians usually produce poor depth with few geometry details but lots of floaters, which hinders the effective modeling of smooth radiance fields. To address this issue, we propose to learn a radiance field within a continuous view-dependent color field parameterized by neural networks. With continuous constraints in color, the bias on low-frequency signals of neural networks can highly encourage the geometry modeling to contribute to high-frequency color variations on images, getting rid of solely relying on the color attribute itself. Moreover, we impose a constraint to improve the Lipschitz continuity of the neural color field, making large changes of each Gaussian's color also match with large changes of its position in terms of a ratio. To this end, we additionally introduce novel self-learning strategies to improve the learning of the neural view-dependent color field without any external knowledge or priors, aiming for better generalization on unseen scenes. Our numerical and visual comparisons on widely used benchmarks justify our idea and show better ability of high fidelity geometry recovery in 3DGS. Project page: https://github.com/yangyixi21/Neural-Color-Fields-3DGS.
Rocco Mattia Di Mauro, Marco Grassia, Giuseppe Mangioni
Graph neural network explanations are often evaluated via deletion fidelity, i.e., by deleting their top-ranked edges and measuring the resulting change in model output. Yet, in a message-passing GNN, "deleting an edge" does not specify a unique intervention. Structural deletion removes the edge from the graph and recomputes quantities such as degree normalization and attention, whereas message suppression leaves this structural context intact while blocking the message carried by the edge. The meaning of deletion fidelity therefore depends on what the intervention targets in the model computation. We formalize this distinction and derive sufficient conditions for exact equivalence. The two interventions coincide for additive message passing with edge-set-independent coefficients, but can diverge when aggregation depends on the presence of zero messages or when removing an edge changes the coefficients applied to surviving messages. We then construct intervention-specific edge attributions and evaluate each resulting ranking under both interventions, testing whether their relative fidelity changes with the evaluation intervention. Across nine datasets and a panel of GIN, GCN, and GATv2 models at depths 2--5, all nine GIN controls satisfy the equivalence predicted by the theory. On the target logit, the relative fidelity of the two rankings depends on the evaluation intervention in 10 of 11 reportable GCN and GATv2 cells, and in nine of them each ranking has the larger deletion area under the intervention it was built with. Deletion fidelity is thus not a property of an edge ranking alone. Evaluation protocols should explicitly report the intervention, output functional, target, and perturbation unit, and test alternative intervention semantics when the scientific question does not uniquely determine one.
Practitioners who compare systems with an LLM judge and have no human labels often check a conclusion by repeating it with a judge from another provider. We test how far that check can be trusted. In a preregistered study (with a disclosed post-registration reduction of the judge set), five judge models from three providers re-judged 4,000 Chatbot Arena battles (nine systems, 33 system pairs) that carry human votes, each in both response orders. Agreement with the human vote was low at the verdict level (median Cohen's kappa = 0.22), and system rankings did not survive a change of judge: Kendall's tau with the human ranking ranged from 0.17 to 0.89 (median 0.67), so the preregistered hypothesis that rankings remain stable was not supported. Conclusions replicated by two low-cost judges from different providers matched the human conclusion more often than single-judge conclusions (+0.10 precision), but the gain came from filtering out one poorly aligned judge, and replication involving that judge left a conclusion for only 35% of comparisons. Judges that agreed with humans reached a stable conclusion with no more verdicts than human votes required; the others failed to converge on 26-70% of system pairs at the sample sizes tested (up to 100 verdicts per pair). Disagreements with the human vote co-occurred across judges far more often than independence predicts, although a single human vote per battle cannot separate shared judge bias from voter noise.
Kaihong Ouyang, Zihao Su, Chengxuan Zheng, Weiyi Lai, Dong Zhou
Long-horizon robotic assembly requires reliable perception and action across grasping, alignment, and insertion. Vision captures global object pose and scene context, where as fingertip touch provides local contact cues that are especially useful as contact states change. We present a visuotactile diffusion policy that combines four-step histories from two camera views and two fingertip tactile images to condition action-sequence generation. A recurrent temporal representation integrates recent observations from both modalities. To encourage it to capture how contact evolves under action, an auxiliary objective predicts fingertip-image changes two steps ahead from the temporal representation and demonstrated actions. This objective is used only during training and adds no inference-time computation. Across three independent training seeds in MuJoCo assembly experiments, the full model achieves an average success rate of $59.0\% \pm 2.6\%$, improving on the visual-only baseline ($42.7\% \pm 1.5\%$) by 16.3 percentage points.
We reassess five economic claims about the cryptocurrency trend factor (CTREND): cross-sectional return predictability, high risk-adjusted performance, incremental information beyond market, size, and momentum, the ability to replace momentum in factor models, and robustness among large and liquid coins and after trading costs. Our reconstruction and code audit identify issues in information timing, technical indicators, forecast coverage, portfolio formation, and return accounting. We first correct future-information use and related settlement rules, then implement broader corrections to indicators, forecast availability, and portfolio accounting. Without additional coin exclusions, the first stage changes CTREND's mean weekly gross return from 4.07% to 4.25%, annualized Sharpe ratio from 2.08 to 1.48, and weekly alpha relative to published market, size, and momentum factors from 2.77% to 3.06%. After comprehensive correction, these become 0.05%, 0.01, and -1.38%; the alpha is no longer significant. Replacing momentum with CTREND yields an advantage in maximum squared Sharpe ratios of 0.058 in the original implementation, 0.024 after the first-stage correction, and -0.0185 after comprehensive correction, with a statistically significant overall decline. Mean weekly gross returns for the largest 100 coins change from 2.93% to 3.99% to 1.00%, while mean weekly net returns for the full long-short portfolio under the lowest assumed trading costs change from 2.93% to 3.35% to -0.87%. These findings show that reported performance and factor-substitution gains depend materially on implementation and sample composition. We distinguish predictive information, an independent return premium, and the incremental value of replacing an existing factor as separate economic claims.
Large audio language models (LALMs) can process speech directly, but it remains unclear how well they use the non-lexical acoustic information that speech carries. We study this question through pain, using TAME-Pain, which contains pain-neutral Harvard sentences and spoken pain ratings recorded at different self-reported pain levels. We introduce HEAR, a four-stage diagnostic framework that traces pain-related information from the speech signal to the model's prediction, and apply it to seven closed-source and open-weight LALMs. On the Harvard sentences, where pain can only be inferred from the voice, all models perform at chance level, and same-speaker references or few-shot examples improve balanced accuracy by no more than six points. On the spoken ratings, the same models reach 87.5-99.7% balanced accuracy. Layer-wise probing of open-weight LALMs shows that pain can be decoded from early audio representations (74.1% balanced accuracy for binary prediction and 45.0% for three-class severity), but that this information, along with most pain-associated acoustic cues, decreases in deeper layers. In an acoustic perception test, weaker models cannot distinguish even large changes in these cues, whereas stronger models can but still fail to infer pain. Finally, we recover pain-related acoustic cues and directly reinject them as structured text into the prompt, which improves pairwise pain prediction by up to 32 points. These results indicate that current LALMs encode paralinguistic information in early layers that is largely unavailable when the model produces its answer.
Pretraining and alignment update the same transformer weights, and we ask in which directions each stage changes them. The residual stream gives every matrix a shared coordinate system, in which a weight change can be compared with the principal subspace of the activations the matrix reads and with the prediction subspace of the unembedding. In this frame, alignment updates to the query and key matrices are enriched in the principal input subspace, while its updates to the output matrices are close to random relative to the prediction subspace, and the same signature appears in late pretraining and under every objective we apply to one checkpoint, so it is a property of the training stage rather than of the objective. The gradient of a linear map is a sum of outer products $δ_t a_t^\top$, and the residual stream enters this product through the activation $a_t$ for matrices that read from it and through the upstream gradient $δ_t$ for matrices that write into it. Read gradients therefore inherit the spectrum of the activations, which is the same under every objective, while write gradients inherit the geometry of the upstream gradients, which point into the prediction subspace only early in pretraining, when the output matrices acquire their prediction geometry. The raw read gradient follows the activation spectrum, the optimiser attenuates the concentration it inherits, and cancellation between tokens sets the scale of the update but not its direction. A write direction derived from the unembedding still changes predictions after alignment, and adaptation restricted to any single group of matrices remains effective, so the signature describes where training places its updates, not which matrices can carry them.
Donghyeon Cho, Joowon Kim, Seungho Shin, Eunho Yang
Diffusion vision-language models (DVLMs) process their entire input at every denoising step. Since visual tokens make up 89-97% of this input on the benchmarks we study, most of the decoding computation is spent repeatedly processing the same image. Visual token pruning methods partially relieve this burden by removing part of the visual tokens, but they lose the removed information and still process the remaining image at every step. To measure how often the image is actually needed, we compare LLaDA-V's predictions with and without the image at every decoding step and find that removing the image changes a token being unmasked at only 22.8% of steps on average and at as few as 7.9%. We therefore propose LOWN, a decoding method that decides at every step whether to process the image and how many tokens to unmask. LOWN predicts the masked tokens without the image and processes the image only when this prediction disagrees with the last image-conditioned prediction. Agreement certifies the earlier prediction for reuse, and the number of certified tokens sets how many are unmasked in that step. We further observe that the need for the image differs across benchmarks, so LOWN's adaptive decision is needed rather than a fixed schedule. Without training or a tuned threshold, LOWN retains 99.6% of LLaDA-V's score on ten benchmarks at 35.3% of the decoding FLOPs on average. LOWN also generalizes across LLaDA-based DVLMs, composes orthogonally with visual token pruning, and runs under key-value caching for further efficiency gains. Code and project page are available at https://hyeon9698.github.io/LOWN/
Daily item co-occurrence embeddings are paired with next-day behavior without future-behavior leakage. Multi-day click-through rate (CTR) training therefore mixes versions whose directions and norms may differ for the same item. We compare Word2Vec and an item-graph XSimGCL adaptation across initialization, inheritance, post-processing, training constraints, and downstream augmentation. The inventory contains 48 configurations, including 32 common-protocol CTR evaluations. Observed metrics are reported as relative percentage changes from the corresponding Word2Vec baseline (0%), with raw values and exact dataset sizes withheld. Three findings emerge. First, multiple upstream interventions improve compatibility and CTR. ID-init and orthogonal alignment increase cosine by 51.2% and 93.4%, and GAUC by 2.2% and 2.6%, relative to ordinary Word2Vec. Retrieval and geometry remain insufficient downstream surrogates. Second, warm starts perform strongly but accumulate scale. The day-matched norm increase over Word2Vec reaches 102.5% after eight days; daily global rescaling limits it to 8.2% and attains the highest observed GAUC increase, 4.0%. Third, CTR learning through a dimensionality-reducing encoder further adapts to residual drift. Against dimension-matched random-projection, SVD, and untrained-network controls, trained encoders attain the highest cross-day cosine for four of six upstream methods. The other two rank third and second in train-to-test cosine. These industrial results support complementary roles for upstream stabilization, inherited-scale management, and downstream task learning.
Individual-level genotype data underpin genetic discovery, disease-risk modeling, and therapeutic development, yet many critical settings remain data-limited, including underrepresented ancestries, rare or imbalanced disease cohorts, and restricted-access biobanks. Conditional synthetic genome generation would support privacy-aware benchmarking, simulation, and data augmentation, but requires learning an ultra-high-dimensional discrete distribution over hundreds of thousands to tens of millions of correlated single-nucleotide polymorphisms (SNPs), or sites of DNA variation. We introduce HELICS: High-fidelity gEneration via Latent Interpolant Conditional Synthesis, a phasing-free latent flow matching framework for genome-wide SNP-level generation from biobank-scale genotype data. HELICS tokenizes chromosomes with convolutional autoencoders, reducing >600K SNPs to 4,096 continuous latent tokens, and trains a transformer parameterized conditional flow matching model over the concatenated multi-chromosome latent representation. Trained on over 417K genomes in the UK Biobank, HELICS generates genome-wide common-variant synthetic cohorts without haplotype phasing and achieves stronger fidelity to real genotypes than existing haplotype-based simulators, including high allele-frequency agreement in external evaluation and superior preservation of local and long-range SNP correlation structure. Conditioning on ancestry and genetic liability summaries enables targeted generation across genetic risk profiles; in perturbation analyses, by increasing the conditioned disease trait liability, HELICS induces SNP-level dosage changes in generated genomes that are highly correlated with published association weights. HELICS provides a scalable generative modeling framework for SNP-level genome data and a path toward conditional synthetic cohorts for data-limited genetic studies.
Ge Chang, Jinbo Su Renmin University of China, Jiacheng Liu, Pengfei Yang, Yuhao Shang, Huiwen Zheng, Hongli Ma GDS, Yan Liang 等 10 位作者
Integrating textual graphs into Large Language Models (LLMs) is promising for complex graph-based QA. However, a key bottleneck is retrieving informative yet compact subgraphs that fit the LLM context. Existing retrievers often struggle, relying either on shallow embedding similarity or costly interactive policies that require excessive supervision.To address these challenges, we introduce Graph-S^3, an agentic textual graph reasoning framework featuring an LLM-based retriever trained with synthetic stepwise supervision. Rather than relying on final answer rewards—which often yield sparse and unstable signals—we optimize the retriever by evaluating each step against offline-extracted golden subgraphs.Our approach distills golden subgraphs via a specialized data synthesis pipeline to formulate dense rewards, facilitating a two-stage training scheme that effectively learns the interactive graph exploration policy.Based on extensive experiments on three common datasets in comparison with seven strong baselines, our approach achieves an average improvement of 15.6% in accuracy and 17.2% in F_1 score. The advantage is even higher in more complicated multi-hop reasoning tasks.
尚无 AI 解读，展示原文摘要
cs.CL · cs.AI
Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表
Zihan Gu Institute of Information Engineering, Chinese Academy of Sciences, TianYi Zhang University of the Chinese Academy of Sciences, Xinyan Zhang independent, Zhiyuan Wang, Han Zhang, Yuhao Wei, Jiacheng Lu Shanghai Jiao Tong University, Tianyi Ma 等 11 位作者
Model editing provides a promising mechanism for updating large language models (LLMs) without expensive retraining. Existing approaches, particularly locate-and-edit methods based on least-squares optimization, aim to introduce targeted knowledge changes while preserving pre-trained behavior. In this work, we show that this objective is fundamentally fragile under standard single-edit evaluation protocols. We first develop a unified theoretical framework that characterizes activation-based editing as a constrained intervention on intermediate representations. Within this framework, we demonstrate that least-squares edits cannot, in general, isolate target updates from unrelated activations, giving rise to unavoidable interference that accumulates with successive edits. Crucially, this degradation can remain undetected in single-edit settings when assessed using conventional success and locality metrics. To expose such hidden instabilities, we introduce an uncertainty-based evaluation protocol that combines structured semantic perturbations with uncertainty quantification based on Sampling with Perturbation for UQ. By measuring edit-induced growth in aleatoric and epistemic uncertainty, our method reveals local knowledge conflicts that are invisible to existing benchmarks. Extensive experiments across multiple models, datasets, and editing algorithms show that both least-squares and other parameter-update-based methods consistently increase post-edit uncertainty. Together, our results suggest that current evaluation practices substantially overestimate the reliability of single-edit model editing, and that uncertainty-based diagnostics are necessary for assessing edit stability.
尚无 AI 解读，展示原文摘要
cs.CL
Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表
Yi Jing, Weiyun Qiu nanjing university, Yihang Peng, Zhifang Sui Peking University
Language change both reflects and shapes social processes, and the semantic evolution of foundational concepts provides a measurable trace of historical and social transformation. Despite recent advances in diachronic semantics and discourse analysis, existing computational approaches often (i) concentrate on a single concept or a single corpus, making findings difficult to compare across heterogeneous sources, and (ii) remain confined to surface lexical evidence, offering insufficient computational and interpretive granularity when concepts are expressed implicitly. We propose HistLens, a unified, SAE-based framework for multi-concept, multi-corpus conceptual-history analysis. The framework decomposes concept representations into interpretable features and tracks their activation dynamics over time and across sources, yielding comparable conceptual trajectories within a shared coordinate system. Experiments on long-span press corpora show that HistLens supports cross-concept, cross-corpus computation of patterns of idea evolution and enables implicit concept computation. By bridging conceptual modeling with interpretive needs, HistLens broadens the analytical perspectives and methodological repertoire available to social science and the humanities for diachronic text analysis.
尚无 AI 解读，展示原文摘要
cs.CL
Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表
Han, Leezy, Kim, Seunggyu, Shim, Dongseok, Lee, Hyeonbeom
Monocular depth estimation (MDE) has been widely adopted in the perception systems of autonomous vehicles and mobile robots. However, existing approaches often struggle to maintain temporal consistency in depth estimation across consecutive frames. This inconsistency not only causes jitter but can also lead to estimation failures when the depth range changes abruptly. To address these challenges, this paper proposes a consistency-aware monocular depth estimation framework that leverages wheel odometry from a mobile robot to achieve stable and coherent depth predictions over time. Specifically, we estimate camera pose and sparse depth from triangulation using optical flow between consecutive frames. The sparse depth estimates are used to update a recursive Bayesian estimate of the metric scale, which is then applied to rescale the relative depth predicted by a pre-trained depth estimation foundation model. The proposed method is evaluated on the KITTI, TartanAir, MS2, and our own dataset, demonstrating robust and accurate depth estimation performance.
尚无 AI 解读，展示原文摘要
cs.CV
Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition · 2026 · 正式发表