Assessing memorization in LLM code generation is challenging because code reuse is not inherently harmful. Correct programs often share APIs, templates, and algorithmic idioms, so code similarity alone may reflect benign reuse rather than problematic memorization. We focus on the harmful case: a model fails to adapt when task semantics change and instead produces code that remains highly similar to the original solution. We propose \emph{code rewriting}, an automated semantic perturbation pipeline that constructs superficially similar but semantically different variants of existing coding tasks by modifying a core logic element of the reference solution and regenerating the corresponding prompt. To summarize this behavior, we introduce the \emph{Harmful Memorization Index} (HMI), an operational diagnostic that combines similarity to the original solution with performance degradation on rewritten tasks. We use harmful memorization as an operational behavioral term for this specific failure mode; HMI does not establish whether an original task or solution appeared in a model's training data. HMI is intended to complement standard accuracy metrics by capturing brittleness under small semantic changes. Across MBPP+ and BigCodeBench, we find that: (1) HMI does not monotonically increase with model scale and often decreases on simpler tasks; (2) high-HMI failures are concentrated in specific logic-edit types, especially constraint changes and data transformations; (3) supervised fine-tuning improves original-task accuracy but can substantially increase HMI on held-out tasks; and (4) across the evaluated post-training settings, PPO yields a more favorable accuracy--HMI trade-off than SFT.
Nan Huang, Mario Tapia-Pacheco, Kun Zhou, Yiming Huang, Kevin José Barrientos Díaz, Tiffany Amariuta, Jingbo Shang
Reliable automated research requires agents to vet data, verify analyses, and generate hypotheses grounded in trustworthy evidence, potentially reducing routine scientific workload while allowing scientists to focus on interpretation and discovery. Existing benchmarks often only assess analytical task completion or hypothesis generation separately rather than testing whether reliable evidence supports valid and novel claims. We introduce DISCERN (Data Integrity and Scientific Capability: Evidence, Reasoning, and Novelty), a controlled benchmark on real, publicly available datasets that evaluates three key levels of an automated research workflow. The first two levels test data integrity and analysis verification under confounds and tool traps, while the third tests hypothesis generation and revision under adversarial review, including counterfactual cases in which evidence consistent with real data and documented scientific phenomena conflicts with established expectations, motivating alternative explanations and testable hypotheses. Across 203 tasks, eight life-science tracks, and eight models, DISCERN shows that strong aggregate performance can mask level-specific weaknesses. Agents earn perfect scores in only 60.8% of Level 1, 34.2% of Level 2, and 0.6% of Level 3 evaluations, with penalties attributed to rejection of sound data, failure to carry recognized limitations into conclusions, and wide variation in hypothesis production. Cross-track rankings by token and code use are substantially more stable than rankings by evidence judgment, suggesting greater consistency in computational effort than in evidence-based reasoning. These profiles identify opportunities for supervised scientific assistance, but current agents do not yet demonstrate reliable autonomous analysis or discovery. Code and data: https://huggingface.co/datasets/discern-bench-anon/discern-benchmark
Letian Peng, Kun Zhou University of California, San Diego, Longfei Yun Facebook, Yupeng Hou University of California, San Diego, Jingbo Shang University of California, San Diego
Role-playing (RP) agents rely on behavioral profiles to act consistently across diverse narrative contexts, yet existing profiles are largely unstructured, non-executable, and weakly validated, leading to brittle agent behavior. We propose Codified Decision Trees (CDT), a data-driven framework that induces an executable and interpretable decision structure from large-scale narrative data. CDT represents behavioral profiles as a tree of conditional rules, where internal nodes correspond to validated scene conditions and leaves encode grounded behavioral statements, enabling deterministic retrieval of context-appropriate rules at execution time. The tree is learned by iteratively inducing candidate scene–action rules, validating them against data, and refining them through hierarchical specialization, yielding profiles that support transparent inspection and principled updates. Across multiple benchmarks, CDT substantially outperforms human-written profiles and prior profile induction methods, indicating that codified and validated behavioral representations lead to more reliable agent grounding.
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
Yuwei Zhang, Wenhao Yu OpenAI, Shangbin Feng University of Washington, Yifan Zhu, Letian Peng, Jayanth Srinivasa Cisco, Gaowen Liu, Jingbo Shang University of California, San Diego
Despite significant advances in large language models (LLMs), their knowledge memorization capabilities remain underexplored, due to the lack of standardized and high-quality testing grounds. In this paper, we introduce a novel, real-world and large-scale knowledge injection benchmark that continuously evolves without human intervention. Specifically, we propose WikiDYK, which leverages recently-added and expert-curated facts from Wikipedia’s “Did You Know...” entries. Each entry is converted into multiple question–answer pairs spanning diverse task formats from easy cloze prompts to complex multi-hop questions. WikiDYK currently contains 12,290 facts and 77,180 questions, and its design allows for seamless extension with future updates from Wikipedia editors. Through extensive experiments using continued pre-training, we reveal a surprising insight: despite their prevalence in modern LLMs, Causal Language Models (CLMs) demonstrate significantly weaker knowledge memorization capabilities compared to Bidirectional Language Models (BiLMs), exhibiting a 23% lower accuracy in terms of reliability. To compensate for the smaller scales of current BiLMs, we introduce a modular collaborative framework utilizing ensembles of BiLMs as external knowledge repositories to integrate with LLMs. Experiment shows that this framework further improves the reliability accuracy by up to 29.1%. Code: https://github.com/zhang-yu-wei/WikiDYK.
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
Chuhan Wang University of Notre Dame, Xintong Li, Jennifer Yuntong Zhang, Junda Wu University of California, San Diego, Chengkai Huang, Lina Yao University of New South Wales, Julian McAuley University of California, San Diego, University of California, San Diego, Jingbo Shang University of California, San Diego
Multimodal large language models often struggle with faithful reasoning in complex visual scenes, where intricate entities and relations require precise visual grounding at each step. This reasoning unfaithfulness frequently manifests as hallucinated entities, mis-grounded relations, skipped steps, and over-specified reasoning. Existing preference-based approaches, typically relying on textual perturbations or answer-conditioned rationales, fail to address this challenge as they allow models to exploit language priors to bypass visual grounding. To address this, we propose SceneAlign, a framework that leverages scene graphs as structured visual information to perform controllable structural interventions. By identifying reasoning-critical nodes and perturbing them through four targeted strategies that mimic typical grounding failures, SceneAlign constructs hard negative rationales that remain linguistically plausible but are grounded in inaccurate visual facts. These contrastive pairs are used in Direct Preference Optimization to steer models toward fine-grained, structure-faithful reasoning. Across seven visual reasoning benchmarks, SceneAlign consistently improves answer accuracy and reasoning faithfulness, highlighting the effectiveness of grounding-aware alignment for multimodal reasoning.
尚无 AI 解读，展示原文摘要
cs.CL · cs.CV
Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表