Artificial Intelligence · Cross-listed from cs.LG
SoFT: Soft Targets for Generalizable LLM Fine-Tuning
Huihao Jing, Wenbin Hu, Shaojin Chen et al. (+7 more)
Distillation enables student
language models to acquire new capabilities from expert teachers. However, integrating knowledge from multi-teacher, multi-domain demonstrations into a single student remains challenging. We study supervised fine-tuning (SFT) in this setting, where students must acquire diverse capabilities while maintaining generalization beyond the training tasks. Our experiments reveal varying trade-offs between in-distribution learning and out-of-distribution generalization across SFT methods, motivating more explicit control over this balance. To this end, we propose soft-target fine-tuning (SoFT) to balance learning from teacher demonstrations with retaining the Base model's existing capabilities. SoFT sets a minimum target probability for each demonstrated
token while making the smallest KL change to the Base distribution. The resulting objective couples learning from demonstrations with adaptively weighted regularization toward the Base model. We further use domain-specific gradient budgets to control this balance and determine a probability threshold for each trajectory. Experiments on mixed-domain
reasoning and agentic tasks show that SoFT achieves the best overall performance among the compared methods, with improvements in both in-distribution capability acquisition and out-of-distribution generalization.
Read the full abstract ▾
All authors Huihao Jing, Wenbin Hu, Shaojin Chen, Haochen Shi, Zhongwei Xie, Guijia Zhang, Yuxuan Liu, Haoyu Huang, Haoran Li, Yangqiu Song
Artificial Intelligence
Beyond Prompt or Skill? Attribution-Guided Optimization of Modular LLM Programs
Haoran Shou, Haoyue Liu, Yu Huo et al. (+2 more)
Large
language models can solve increasingly diverse
reasoning tasks, yet their performance remains highly sensitive to task
prompts, intermediate instructions, and the way reusable problem-solving knowledge is incorporated. Existing optimization methods usually focus on only one part of this design space: they either optimize a monolithic
prompt, or separately induce and refine skills from model traces. As a result, they lack a principled mechanism for deciding which component should be updated when failures occur, and they rarely optimize
prompts, skills, and skill-use policies in a unified framework. We propose SPARO (Skill,
Prompt, And Routing Optimization), a framework that jointly optimizes task instructions, reusable skill blocks, and routing rules. It performs controlled counterfactual evaluations, converts examples' effects into a probabilistic responsibility distribution over
prompt, skill, and routing components, samples one component from that distribution, and applies the corresponding targeted mutation. This design moves
language-program optimization beyond global
prompt rewriting toward structured, reusable, and selectively activated task knowledge. Across five
benchmarks and five worker models, SPARO consistently outperforms both
prompt-centered and skill-centered optimization baselines. These results suggest that effective
language-program optimization depends not only on discovering useful task knowledge, but also on deciding where that knowledge should be stored and when it should be activated.
Read the full abstract ▾
All authors Haoran Shou, Haoyue Liu, Yu Huo, Kun Zeng, Xiaoying Tang
Artificial Intelligence · Cross-listed from cs.CL
Explaining Textual Entailment with Lexical Entailments: Using LLMs to Supply Lexical Relations for Formal Proofs
Jorryt de Jong, Stefan Moraca, Ettore Cesari et al. (+1 more)
Large
Language Models (LLMs) are highly capable of natural
language reasoning and appear to store a great deal of lexical knowledge, but it is still unclear how much of this knowledge they actually use when
reasoning, and whether they use it in the right way. On the other hand, logic-based Natural
Language Inference (NLI) systems provide transparent and formally grounded
reasoning, but they need to be supplied with rich lexical knowledge to prove inferences beyond purely logical ones. In this paper, we evaluate whether LLMs can identify all lexical knowledge needed to solve NLI problems and how much this knowledge contributes to proof search in a logic-based NLI system. Our research focuses exclusively on structured lexical entailments (e.g., chinchilla$\sqsubseteq$small animal) as a proxy for structured explanations for NLI problems with an entailment label. First, we curate a
dataset for a new task of explaining sentential entailments with a set of lexical entailments. The
dataset is used to intrinsically evaluate LLMs on generating structured lexical explanations. Then, we use NLI as an extrinsic evaluation in a simple neuro-symbolic setting, assessing whether LLMs can supply sufficient lexical relations to LangPro, a natural-logic
theorem prover for natural
language. The results show that the proposed task remains challenging even for hosted proprietary LLMs, and that their contribution to
theorem proving is moderate: generated relations are often only partially sound and may be tailored to the specific NLI problem rather than representing generally valid lexical knowledge.
Read the full abstract ▾
All authors Jorryt de Jong, Stefan Moraca, Ettore Cesari, Lasha Abzianidze
Artificial Intelligence
RepoMAS: Solving Progressively Specified Tasks with Issue-Driven Multi-Agent Systems
Yuchen Song, Andong Chen, Wenxin Zhu et al. (+2 more)
LLM-based multi-
agent systems (MASs) have shown strong potential for solving complex tasks, but most assume that task requirements are sufficiently specified before execution. In practice, user requests are often incomplete, and additional requirements may only become clear during
reasoning, tool use, or execution. We refer to such problems as progressively specified tasks. To systematically study this setting, we introduce ProgSpec, a
benchmark that evaluates final outputs against requirements explicitly stated in the initial request and additional requirements supported by the available task evidence. We further propose RepoMAS, an issue-driven multi-
agent framework inspired by open-source project management. RepoMAS records newly discovered requirements, conflicts, and failures as structured Issues and uses them to revise the task specification and execution structure during problem solving. Across ProgSpec and five existing
benchmarks, RepoMAS achieves the best performance. Further analyses show that its issue-driven revision and repository maintenance mechanisms consistently contribute to performance. These results highlight the importance of allowing MASs to revise not only how a task is solved, but also revise their explicit representation of task requirements during execution.
Read the full abstract ▾
All authors Yuchen Song, Andong Chen, Wenxin Zhu, Muyun Yang, Tiejun Zhao
Artificial Intelligence
When Helpful Text Hurts: Option-Redirecting Bias in Vision-Language Models
Tam Le Thi Thanh, Hoang Tran Van, Hong-Hanh Nguyen-Le et al. (+1 more)
In tri-modal visual question answering (VQA), auxiliary text is commonly used to complement visual and textual inputs, yet its reliability is often uncontrolled. While prior work studies modality conflicts in general, it remains unclear how different types of unreliable auxiliary text affect answer selection under fixed image-question-option contexts. In this work, we show that the most harmful auxiliary text is not necessarily the most factually incorrect, but the one that aligns with the question while contradicting the image and favoring a specific distractor, leading to systematic redirection of model predictions. To isolate this effect, we introduce the Textual Reliability Ladder, a controlled diagnostic protocol that decomposes auxiliary text along three axes: image consistency, question relevance, and option support. Across multiple
datasets (ScienceQA, VCR, A-OKVQA, Causal-VidQA) and recent VLMs, we find that such distractor-supporting text induces the largest accuracy drops (up to 53.1%) and concentrates errors on specific incorrect options. To mitigate this failure mode, we propose a training-free inference-time intervention that explicitly counteracts this redirection effect via noise-stability steering and dynamic grounding, reducing redirected errors while largely preserving performance under faithful text. Our results highlight that auxiliary-text reliability must be understood at the decision level, rather than solely through factual correctness, and provide a practical pathway toward more robust tri-modal
reasoning.
Read the full abstract ▾
All authors Tam Le Thi Thanh, Hoang Tran Van, Hong-Hanh Nguyen-Le, Thanh Duc Ngo
Artificial Intelligence
Towards Scalable Data Diversification for Language Model Pretraining via Leverage Score Sampling
Zailin Ma, Quzhe Huang, Yujun Li et al. (+2 more)
Data selection for
language model pretraining faces a fundamental tension between quality and diversity. While quality filtering is empirically effective, it often induces diversity collapse: by favoring texts similar to high-quality reference corpora (e.g., educational or QA-style data), it systematically excludes valuable data from underrepresented domains. In contrast, diversified selection preserves domain balance and encourages robust downstream performance, yet existing methods either focus on coverage-oriented objectives that indirectly enhance diversity, or directly optimize for diversity via costly covariance
matrix recomputation that limits scalability. To address these issues, we introduce Leverage Score Sampling (Lev), which iteratively selects samples that maximally expand the determinantal volume of the embedded data via leverage scores, a computationally efficient criterion that eliminates
matrix recomputation and enables scalable selection. Empirically, Lev delivers up to $72\times$ speedup and improves
dataset diversity, measured by the Vendi score, by $9.2\%$ over the strong diversification baseline DiSF. On CommonCrawl (CC) web data selection, Lev improves accuracy across seven downstream tasks by up to $1.31\%$ over existing baselines. For domains where robust quality criteria are inherently difficult to define (e.g., code), Lev serves as an effective unsupervised curation alternative: on StarCoderData, the selected subset reduces bits-per-byte by $3.08\%$ over DiSF. Notably, we uncover a cross-domain collapse of quality filtering: CC data filtered by DCLM-fastText fail to retain sufficient code-related content, yielding inferior code performance relative to Lev-selected data. These findings advocate for integrating diversity-aware practices into quality filtering for more effective data curation in
language model pretraining.
Read the full abstract ▾
All authors Zailin Ma, Quzhe Huang, Yujun Li, Congyuan Rao, Yaodong Yang
Artificial Intelligence
Separating Diagnosis from Disease Representation: Dual-View EEG Learning with Neural-Dynamics-Guided Deformation
Jiaying Wang, Shouqian Shi, Yutong Chen et al. (+5 more)
Electroencephalography (
EEG)-based closed-loop neuromodulation calls for a subject-specific structured state, as opposed to a single disease probability, specifying which
brain regions are deviant, at which frequencies, and at which lags. Sensor-space models keep the strongest diagnostic evidence without anatomy, source-space models give anatomy at a loss of predictive signal, and post-hoc attributions stay outside the prediction. We separate the two instead of forcing them into one representation, and propose DMD-
EEG (Dual-view Multiscale Deformation for
EEG), which keeps a fixed scalp spectral expert for diagnosis and models the source-space disease-related representation as a low-rank, sparse, iterative deformation of a healthy neural-dynamics prior in a $46$-region-of-interest (ROI) $\times$ $5$-frequency $\times$ $4$-lag (autocorrelation-timescale) space. The two experts meet only at a fixed decision level, so the source state is architecturally separate from the scalp expert. Across major depressive disorder (MDD), first-episode psychosis (FEP), and Parkinson's disease (PD), decision-level fusion matches the strongest single expert on MDD and FEP and exceeds the source branch on PD. On FEP the source expert is the strongest branch, the task where the deformation contributes most. The source state is an explicit ROI-frequency-lag attribution defined in a shared source coordinate system across montages, which we treat as an anatomically-coordinated predictive representation whose coordinates are directly readable and hypothesis-generating. The highest-saliency coordinates align with established disease circuitry (fronto-limbic-temporal regions in MDD, motor-cortex beta in PD), and the MDD state transfers by rank to an unseen cohort recorded with a different montage.
Read the full abstract ▾
All authors Jiaying Wang, Shouqian Shi, Yutong Chen, Xu Yang, Jie Chen, Xingyu Pan, Lei Zhang, Sheng Zhong
Artificial Intelligence
From Outcomes to Strategies: Learning Strategy Utility for Mathematical Reasoning
Ruikang Zhang, Xiao An, Xuli Shen et al. (+5 more)
Reinforcement learning with verifiable rewards has substantially improved mathematical
reasoning. However, terminal correctness alone provides limited insight into the quality of high-level strategies, such as
theorem selection and subgoal decomposition, when considered separately from their subsequent execution. This paper studies strategy utility, which is defined as the likelihood that a strategy supports a correct downstream solution under a given executor. We introduce SURE, a framework for learning and leveraging relative strategy utility. In this framework, high-level strategies are separated from their detailed
reasoning. Based on the pairwise preferences constructed from strategy-conditioned rollouts and teacher-generated contrasts, a Strategy Reward Model is learned to estimate relative strategy utility. During reinforcement learning, the frozen reward model reads only the extracted strategy, whose score is combined with the correctness and format rewards in a sequence-level GRPO objective. Compared with outcome-and-format GRPO baselines, experiments show that SURE improves average pass@1 by 1.87%, 2.64%, and 2.93% across three policy backbones. Our method also achieves competitive or better accuracy than stronger reward baselines while requiring substantially lower GRPO-stage compute.
Read the full abstract ▾
All authors Ruikang Zhang, Xiao An, Xuli Shen, Jiaxing Sun, Xiaoyi Yu, Jin Zeng, Jiang Wu, Tong Lin
Artificial Intelligence
VPEvolve: A Self-Evolving Virtual Process Engineer for Computational Lithography
Tianyi Li, Wenxuan Dong, Donger Luo et al. (+5 more)
Optical proximity correction (OPC) recipes grow as engineers add local rules to repair newly discovered lithography hotspots. Each correction can interact with existing rules, while lessons from commercial-tool trials remain scattered across code and logs. \system combines a Virtual Process Engineer (VPE) harness with a Skill Bank of measured engineering experience. The harness equips a frozen
language model with process manuals, layout analysis, recipe editing, and commercial-tool evaluation. The actor proposes changes to the global parameters, local targeted rules, or diagnostic trials. After each evaluation, an LLM reflector and curator turn the measured response into evidence-linked judgments. The actor retrieves them before its next trial. Feasible improvements update the retained recipe; every measured trial informs the Skill Bank that guides the next edit. The model weights remain fixed. On a FreePDK45-derived
benchmark with ten commercial-tool evaluations per case, \system reduces the mean per-case maximum edge placement error from 18.294 to 5.361 nm on Poly and from 22.052 to 15.692 nm on Metal1. Every final recipe satisfies the predefined quality constraints and improves the maximum error by at least 0.1 nm.
Read the full abstract ▾
All authors Tianyi Li, Wenxuan Dong, Donger Luo, Nan Wang, Yanpeng Chen, Jiaqi Liu, Xinyun Zhang, Hao Geng
Artificial Intelligence
PULSE: Identifying Demonstration-Utility Features with Sparse Autoencoders
Chenduo Hao, Chuanbao Gao, Pinjun Zeng et al. (+4 more)
In-context learning is highly sensitive to demonstration choice, yet most methods select demonstrations using external query-demonstration similarity. Such criteria can miss model-specific signals: Similar demonstrations may activate different internal features and downstream behaviors. We introduce PULSE (Paired Utility Localization over Sparse Encodings), an SAE-based framework for identifying model-internal features associated with demonstration utility and using them for demonstration selection. Using a small labeled discovery set, PULSE samples candidate demonstration sets, measures their zero-shot-relative utility under the target model, and scores SAE features by how their activation differences align with utility differences. The top positive and negative coordinates form a sparse utility-localization vector. We use this vector in two complementary ways: as a signed score for controlled complete-set ranking, and as PULSE-Retriever, which converts its magnitude into a feature-relevance mask for scalable pool-scale
retrieval. Across classification, generation, and
reasoning benchmarks, PULSE-Retriever improves over the strongest baseline by 2-3 accuracy points, 0.6-0.9 BLEU-4, and 3.2 exact-match points, respectively, while controlled ranking validates the identified features encode a predictive set-level utility signal. Feature inspection and cross-
dataset experiments suggest that the identified features capture task-relevant,
dataset-conditioned patterns, yet retain utility signals that partially transfer across
datasets. Our code is available at https://github.com/aohenuo/PULSE.
Read the full abstract ▾
All authors Chenduo Hao, Chuanbao Gao, Pinjun Zeng, Jingze Zhu, Chonghan Liu, Zidong Liu, Xu Yang
Artificial Intelligence · Cross-listed from cs.LG
Bison: Cross-Dataset Learning for Unseen-Compound Perturbation Prediction
Yunfan Liu, Kasra Ghorbani, Yufei Huang et al. (+6 more)
Predicting transcriptional responses to unseen compounds is limited by fragmented chemical coverage and heterogeneous experimental platforms and gene panels. To assess molecular generalization across these settings, we build on Chem-PerturBridge to
benchmark eight
datasets with 16,771 compounds, withholding test compounds from every training
dataset. This comparison reveals that high overall response agreement can coexist with weak prediction of drug-specific differences, despite reproducible signals across repeated measurements. To exploit complementary chemical supervision while targeting these differences, we introduce Bison: a shared gene representation connects native panels, while two discrete
diffusion models compose context-dependent responses with molecular deviations learned through matched drug-contrast supervision. A single Bison model jointly trained across all eight
datasets achieves the highest mean overall-response and drug-contrast Pearson correlations on the full
benchmark in comparison with 11 methods trained independently per
dataset. Compared with
dataset-specific training of the same architecture, joint training increases mean drug-contrast correlation by 27.4%, with gains across all eight
datasets and improvements in overall response prediction. These results demonstrate how matched drug contrasts turn complementary screens into shared molecular supervision for unseen-drug response prediction while preserving native gene measurements.
Read the full abstract ▾
All authors Yunfan Liu, Kasra Ghorbani, Yufei Huang, Zicheng Liu, Jiangbin Zheng, Jingbo Zhou, Shaorong Chen, Chang Yu, Stan Z. Li
Artificial Intelligence · Cross-listed from cs.LG
Adapting Nonstationary Multi-output Gaussian Processes to Bayesian Optimization
Zikai Xie
Multi-objective
Bayesian optimization (MOBO) commonly relies on independent Gaussian processes (GPs) with stationary kernels, limiting its ability to represent nonstationary structure and share information between objectives. However, expressive nonstationary GPs do not necessarily make reliable BO decisions. We study this mismatch for the multi-output low-rank nonstationary (MO-LRN) GP: strong training fit can coexist with large off-design errors and optimistic acquisition predictions. We introduce MOLRN-BO, which combines a regularized shared-spectral surrogate with objective-specific residuals, prequential mean correction and tempered covariance scaling, and Pareto-local qLogEHVI optimization with periodic global search. Experiments on 12 deterministic bi-objective
benchmarks show that MOLRN-BO substantially improves upon the original MO-LRN and achieves the best average problem ranks for final normalized hypervolume and normalized inverted generational distance among nine evaluated
algorithms. It also achieves the strongest adverse-tail performance while remaining competitive with the leading baselines in anytime optimization. Ablation studies further show that the shared spectral construction improves off-design prediction, the local–global decision policy improves optimization performance, and hierarchical calibration reduces systematic candidate bias. These results demonstrate that nonstationary multi-output surrogates can deliver strong and robust MOBO performance when their structure and use are explicitly adapted to the demands of sequential optimization.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.CV
Seeing Parts, Reasoning about Worlds: Visual Inference under Partial Observation
Wei Wang, Wenqiao Zhang, Yutong Lin et al. (+2 more)
World modeling under partial observation requires
reasoning about the complete worlds that remain compatible with limited visual evidence. Occluded objects and unseen regions can leave several world states possible; additional views can exclude alternatives and strengthen the conclusions supported by the observations. We introduce WorldScope to study this process through possible-world semantics, evidence-grounded data, and learned visual representations. WorldScope-1.2M provides 1.2 million English question-answer pairs spanning eight world properties, ten task interfaces, and three observation protocols. Its answers encode confirmed facts, supported bounds, and unresolved possibilities. Complementary supervision comprises 4,800 certified counterworld groups with equivalent base observations and different hidden object configurations and query answers. These groups provide physical witnesses of ambiguity and training-only labels for world compatibility and view-induced exclusions. We propose WorldFlow, which composes cross-view entity evidence and support-surface coverage into an image-subset evidence
lattice. Counterworld compatibility and transition objectives train subset representations to reflect how new observations constrain possible worlds. A shared answer generator uses these representations to predict the strongest supported conclusion. WorldScope-Bench evaluates claim judgments and evidence-dependent conclusions as views are selected, combined, removed, or ordered. On its 5,000-question test set, WorldFlow reaches 64.34% exact accuracy, improving over the same backbone trained on QA alone by 24.88 percentage points. It retains 50.43% accuracy on the 3,000 questions from structure-disjoint scenes.
Read the full abstract ▾
All authors Wei Wang, Wenqiao Zhang, Yutong Lin, Jun Xiao, Yueting Zhuang
Artificial Intelligence · Cross-listed from cs.CV
De-biasing Skeleton-based Action Recognition with Convex Hull Adaptive Shift
Mengyuan Liu, Yuhang Wen, Yi Zhang et al. (+4 more)
Skeleton sequences can represent both individual actions and multi-entity interactions, encompassing human bodies, hands, objects, and
robots. Existing approaches to recognize skeleton-based actions and interactions usually adopt a late fusion strategy, which expects individuals are independent and identically distributed to train a robust weight-shared entity encoder. However, observed entity bias in various skeletal data violates this assumption, leading to suboptimal optimization of backbone models that might produce wrong recognition results. This bias arises from the world coordinate system's initial configuration, where the choice of origin often creates bias in representation. To this end, we propose a Convex Hull Adaptive Shift based normalization method to reduce Entity bias (CHASE), improving performance across a variety of skeleton-based action and interaction recognition tasks. To adaptively apply plausible shifts to the input skeletons, we formulate a plug-and-play parameterized network that ensures the relocated world origin lies within the skeleton convex hull, which avoids non-convergence by limiting the search space. To further minimize entity bias, we incorporate an auxiliary objective that leverages pair-wise distribution distances to guide network optimization. To support both single- and multi-entity actions, we propose a sub-entity strategy that offers a consistent formulation for both scenarios. Moreover, CHASE demonstrates compatibility with various intra-skeleton modalities, such as bones and velocities, highlighting its adaptability. Essentially, our method works as a normalization approach to reduce entity bias, enabling subsequent classifiers to achieve improved recognition performance across diverse settings. Extensive experiments on 7
datasets verify our approach by seamlessly integrating with various backbones and significantly boosting their performance.
Read the full abstract ▾
All authors Mengyuan Liu, Yuhang Wen, Yi Zhang, Songtao Wu, Hong Liu, Junsong Yuan, Beichen Ding
Artificial Intelligence · Cross-listed from cs.RO
DRAM: Delta-rule Recurrent Associative Memory for Robot Manipulation Policies
Xinyu Zhao, Yixiang Shan, Tao Yang et al. (+5 more)
Robotic manipulation is inherently history-dependent, yet most pretrained robotic policies condition on only the current observation or a short temporal window. Equipping such policies with long-term
memory remains challenging: existing approaches either feed the backbone multi-frame observation windows, which substantially increase inference cost, or rely on pre-defined semantic features, which limit task generality and may also require the retraining of the backbone to adapt to the
memory. We introduce DRAM (Delta-rule Recurrent Associative
Memory), a plug-and-play
memory module that can be attached to a wide range of pretrained robotic policies, endowing them with long-horizon
memory without architectural modification or backbone retraining, requiring only task-specific post-training of the
memory module and action expert. DRAM maintains a fixed-size associative
memory using gated delta-rule linear
attention, with a modified update that incorporates all
tokens within each frame in parallel. An architecture-agnostic readout integrates historical context into action prediction across different policy architectures. Experiments show that DRAM consistently improves frozen pretrained policies over short-context baselines and alternative compact
memory designs, validating its effectiveness as a fixed-size, post-hoc
memory module trained with the backbone frozen.
Read the full abstract ▾
All authors Xinyu Zhao, Yixiang Shan, Tao Yang, Runyu Lei, Yiming Zhao, Jiaxin Fan, Zongbao Feng, Peng Jia
Artificial Intelligence
ForkLeft: Entropy-First Rollouts for Prefix-Aligned Autoregressive-to-Diffusion Distillation
Junming Liu, Jicheng Wang, Yifeng He et al. (+2 more)
Autoregressive Next-
Token Prediction (NTP) has enabled strong
reasoning capabilities in
language models, while
Diffusion Language Models (DLMs) offer flexible
token orders and parallel generation. We ask whether DLMs can acquire NTP-style
reasoning through
distillation without giving up their native generation process. Direct
distillation, however, faces a fundamental mismatch: an autoregressive teacher predicts from a left prefix, whereas a DLM can condition on
tokens on both sides. We introduce ForkLeft, a
distillation framework that resolves this mismatch by separating the student's rollout from teacher supervision. During training, the student first performs
entropy-first rollouts that commit uncertain positions and expose potential forks. We then fix the resulting student prefix and distill an NTP teacher under the same context, with answer correctness determining the supervision source. At inference, the student returns to its native confidence-first parallel decoding. With Qwen3-30B-A3B-Base, ForkLeft improves Efficient-DLM-4B on all ten
benchmarks, raising MATH500 from 72.60% to 79.60% and consistently outperforming three alternative designs. The gains scale with teacher strength and generalize to SDAR-4B with only $500$ updates. At matched scale, the distilled 4B and 8B students exceed the published SDAR-Chat and OPDLM models on seven
benchmarks, showing that DLMs can learn NTP-style
reasoning without sacrificing native parallel generation. Code and
datasets will be released upon acceptance.
Read the full abstract ▾
All authors Junming Liu, Jicheng Wang, Yifeng He, Hao Chen, Jianzhong Qi
Artificial Intelligence · Cross-listed from cs.LG
Rethinking Training-Inference Mismatch in LLM Reinforcement Learning: Where It Arises and How to Correct It
Tianrun Yu, Kaixiang Zhao, Shangzhe Li et al. (+4 more)
We study training-inference mismatch in reinforcement learning with verifiable rewards (RLVR) for large
language models, where rollouts are sampled by an inference engine while gradients are computed by a training engine, and the two engines assign different probabilities to the same
tokens. To account for this discrepancy in policy updates, we introduce calibrated importance sampling (CIS). CIS is motivated by an empirically supported logit-displacement characterization that expresses the mismatch as an additive displacement $\varepsilon_t$ in log-odds, determined by the per-logit perturbation before the softmax, whose distribution is approximately invariant to
token confidence. This characterization motivates a confidence-aware truncation: large positive displacements are truncated at a single constant threshold, which maps back to an importance-ratio cap that tightens as
token confidence increases. Theoretically, we show that CIS replaces the unbounded second moment that governs the error of exact importance sampling with a term bounded by a constant, at the cost of a bias controlled by the truncated excess. In evaluation across three mixture-of-experts models and five mathematical
reasoning benchmarks, CIS achieves the highest five-
benchmark average on all three models among the evaluated baselines. Diagnostic analyses show that CIS places less truncation bias on low-confidence
tokens than truncated importance sampling, while upward clipping of small importance weights reduces held-out accuracy.
Read the full abstract ▾
All authors Tianrun Yu, Kaixiang Zhao, Shangzhe Li, Yuxiao Yang, Porter Jenkins, Weitong Zhang, Taylor W. Killian
Artificial Intelligence
Controllable GNN Explanations via Multi-Metric Preference Selection
Rachit Verma, Yashraj J. Deshmukh and Anirban Dasgupta
Mechanisms for generating GNN explanations are crucial for building trust and mitigating biases in Graph Neural Networks (GNNs), especially in high-stakes scenarios. Most current methods optimize only for fidelity under the sparsity constraint. However, this discounts the need for interpretable explanations (those that consist of familiar motif patterns) and stable explanations (those that remain unchanged under structural perturbations). We propose a novel approach that optimizes GNN explanations across these metrics, exposing their relative weighing as a control. Experiments on various real-world
datasets, including MUTAG, BA-2Motif, BAMultiShapes, and
PROTEINS, suggest that our method produces higher-fidelity explanations than a state-of-the-art baseline on MUTAG and
PROTEINS across all evaluated budgets, and on BA-2Motif at larger budgets, while being faster in the regime of small explanation budgets. They also indicate that our method produces more stable explanations for the MUTAG and BA-2Motif
datasets. We also examine the relationship among different metrics via their induced trade-off surface and explore its dependence on the nature of the motif library.
Read the full abstract ▾
Artificial Intelligence
From Latents to Wires: Surgical Post-Editing on Large Language Models
Jiankai Jin, Xiangzheng Zhang, Zhao Liu et al. (+4 more)
Given a large
language model (LLM), can whoever holds the weights name a semantic target (e.g., the model's identity), locate the model components that produce it, and edit them so that the target no longer appears while other capability is preserved? We call such an edit on a trained model a post-edit. We present L2W (latents to wires), a framework that performs
surgical post-edits for named semantic targets. For localization, L2W uses Jacobian lens (J-lens) attribution to score components against the semantic target. For
surgical removal, because LLM mechanisms are redundant (i.e., a semantic target may have multiple components producing it), L2W runs Counterexample-Guided Causal Cut (CGCC) until the target no longer appears. CGCC first cumulatively closes model components, treating each surviving expression of the target as a counterexample that exposes the next components to close, and then reopens some of them to preserve capability. In a controlled experiment with an implanted behavioural watermark, L2W removes the watermark, and its localization lands on the model region the implant changed. Across three model configurations, L2W removes model-metadata (e.g., identity) self-claims in all nine runs, and adult-content refusal in all three, with no held-out target residual. L2W further composes two post-edits on a text-to-image model: one removes the refusal of requested nudity, and a second removes the nude rendering the first exposes. The results support post-editing as a complement to post-training: post-training installs preferred behaviours, and post-editing removes named unwanted ones.
Read the full abstract ▾
All authors Jiankai Jin, Xiangzheng Zhang, Zhao Liu, Wenzhuo Xu, Dongdong Yang, Deyue Zhang, Quanchen Zou
Artificial Intelligence
Multi-Agent System Search via Active Substructure-aware Policy Optimization
Beicheng Xu, Bowen Fan, Weitong Qian et al. (+2 more)
LLMs enable multi-
agent systems (MAS) to tackle complex tasks, but manually designing
agent roles,
prompts, and communication structures requires substantial expertise and effort. This motivates learning policies that construct query-specific MAS from execution reward. Existing approaches typically train these policies by repeatedly traversing a fixed set of training queries and assigning rewards at the workflow level. However, this overlooks differences in queries' evolving learning potential and obscures which substructures improve solution quality. In this paper, we propose Active Substructure-aware Policy Optimization (ASPO), a RL framework for query-level MAS search. ASPO introduces an Adaptive Query-Selection Mechanism (AQSM) that focuses training on queries at the policy's competence boundary: those it can solve but not yet reliably. A complementary discovery mechanism widens architectural exploration for hard queries, helping distinguish insufficient exploration from operator capability limits. Beyond query selection, ASPO introduces substructure-level rewards that measure output-quality gains within each action's descendant subgraph. These rewards guide proximal policy optimization to reinforce useful architectural refinements and discourage redundant or harmful computation. Together, these mechanisms prioritize learnable queries and provide fine-grained feedback for learning effective MASs. Across six
benchmarks spanning mathematical
reasoning, general question answering, and code generation, ASPO ranks first on every
benchmark against twelve baselines.
Read the full abstract ▾
All authors Beicheng Xu, Bowen Fan, Weitong Qian, Lingching Tung, Bin Cui
Artificial Intelligence
PrismQuant: Optimal Null-Space Rotations for Grouped Quantizers
Yanlong Chen, Yining Chen, Song Zhang et al. (+2 more)
Smaller activation outliers do not necessarily imply better low-bit quantization: their
alignment with the quantizer matters. We introduce PrismQuant, a quantizer-aware rotation framework that aligns the leading activation eigenspace with the constant group subspace of asymmetric grouped INT4. The affine offsets represent the energy in this subspace without widening the range within the group. We formulate rotation design as a Ky Fan trace maximization and derive a closed-form solution that is provably optimal for this
alignment objective. Compact Householder transformations and their compact-WY representation enable gradient-free construction and efficient application at both foldable and online sites. A predictive range law further connects unaligned activation energy and group size to quantization-relevant variation. Experiments on Llama, Qwen, and Mistral span dense models up to 70B parameters and a 30B mixture-of-experts model. Under W4A4KV4, PrismQuant sets the state of the art on Llama-3.2-3B among the compared methods in both perplexity and accuracy. On Llama-3.1-70B, it attains 3.85 perplexity and 72.46% average zero-shot accuracy, only 0.22 percentage points below full precision. In the deployment study on Llama-3.1-8B, our optimized implementation achieves 1.51x prefill and 1.22x CUDA Graph decode speedups over matched FP16 baselines, with 56.34% lower decode peak
memory and only 2.35% additional Graph decode latency over Hadamard. Code and models are available at https://github.com/ForeverBlue816/PrismQuant and https://huggingface.co/collections/ForeverBlue/prsimquant.
Read the full abstract ▾
All authors Yanlong Chen, Yining Chen, Song Zhang, Amirhossein Habibian, Yawei Li
Artificial Intelligence
Authorization Closure Graph: Minimal Repair for LLM Agents with Evolving User Instructions
Qingzhuo Wang, CaiYi Wang, Jinglu Meng et al. (+4 more)
Tool-using large
language model (LLM)
agents increasingly perform state-changing actions that require user authorization. Yet existing approaches do not provide a principled mechanism for selectively updating prior authorization when only part of an instruction changes. To this end, we propose an Authorization-Closure-Graph (ACG)-based framework that represents authorization and its dependencies as an evolving, versioned state. ACG selectively invalidates authority affected by a revision while preserving unaffected portions of the authorization state, and computes a minimal repair that identifies only the missing evidence or authority required for execution. This enables
agents to adapt to revised instructions while avoiding stale authority and unnecessary authorization requests. We evaluate ACG across three advanced LLMs in two natural tasks, and ACG consistently improves action safety rate and task success rate. Code is available at https://github.com/weiliang822/ACG.
Read the full abstract ▾
All authors Qingzhuo Wang, CaiYi Wang, Jinglu Meng, Ruiyang Qin, Kunyu Peng, Zhihua Wei, Wen Shen
Artificial Intelligence · Cross-listed from cs.CR
CyberClear: A Benchmark for LLM Agent Systems on APT Attack Chain Provenance
Qi Chen, Fushuo Huo, Hangli Shen et al. (+3 more)
Large
language model
agents have demonstrated promising capabilities in cybersecurity tasks, yet their ability to reconstruct complete Advanced Persistent Threat attack campaigns from complex security logs remains largely unexplored. Existing cybersecurity
benchmarks for
agents mainly focus on vulnerability discovery, exploitation, and security analysis tasks, leaving the evaluation of attack chain provenance under realistic security logs insufficiently studied. To address this gap, we introduce CyberClear, a
benchmark for evaluating LLM
agents and advanced
agent systems on APT attack chain provenance from long-context security logs. CyberClear covers both single-step attacks and multi-stage attack chains, requiring
agents to identify attack evidence, infer attack progression, and generate provenance graphs containing entities, causal relationships, MITRE ATT&CK techniques, and forensic evidence. To enable comprehensive evaluation, we develop an evaluation method tailored to APT attack chain provenance. Unlike conventional text similarity metrics that focus on surface-level matching, our evaluation examines whether reconstructed graphs preserve the semantics of attack chains across single-step behavior correctness, multi-step behavior identification, temporal and causal consistency, entity and relationship fidelity, and overall attack narrative consistency. Advanced multi-
agent systems powered by state-of-the-art LLMs still struggle on CyberClear, motivating us to propose CyberProvenance, an
agent cyber harness designed for multi-
agents that augments LLM
agents with evidence accumulation, execution-based validation, and feedback-guided refinement mechanisms for reliable attack-chain provenance. Extensive evaluations on CyberClear demonstrate the effectiveness of CyberProvenance in improving evidence
reasoning, execution-grounded validation, and complete APT attack chain reconstruction.
Read the full abstract ▾
All authors Qi Chen, Fushuo Huo, Hangli Shen, Jingcai Guo, Shuhao Li, Guang Cheng
Artificial Intelligence
PluginRSI: Recursive Improvement of Agent Harnesses with Reusable Plugins
Yaorui Shi, Yuchun Miao, Yuxin Chen et al. (+5 more)
The harness surrounding a
language model is a central determinant of
agent performance. Recent methods optimize harnesses by searching over complete programs, where individual mechanisms are difficult to isolate and reuse. We introduce PluginRSI, which represents a harness as a composition of atomized plugins and organizes harness
evolution around these plugins. Individual plugins are improved independently and accumulated in a shared library, then recombined into new harnesses at each iteration. PluginRSI improves over existing harness optimization methods across software engineering, command-line interaction, and question-answering tasks. The resulting harnesses retain their advantage when transferred to other solver models without further optimization. The evolved plugin library accelerates subsequent optimization from the initial harness, which helps faster and higher convergence on unseen tasks. These results show that accumulating reusable mechanisms provides an effective basis for continued harness improvement.
Read the full abstract ▾
All authors Yaorui Shi, Yuchun Miao, Yuxin Chen, Jiayuan Zhang, Yueqing Sun, Xierui Song, Xiang Wang, An Zhang
Artificial Intelligence
MergeHEIR: Mitigating Multimodal Hallucinations as the Tax of Model Merging
Jinyu Li, Hao Fang, Zhiming Zhang et al. (+3 more)
Model merging consolidates task-specialized experts into a single deployable model. However, we show that such capability consolidation incurs a merging tax of increased
hallucination: across 8 model-merging methods, every merged checkpoint exhibits a higher
hallucination rate than the average of its constituent experts. An intuitive approach is to adapt existing
hallucination-mitigation methods to the post-merge model, yet this unconstrained adaptation disrupts inherited capabilities, creating a tension between
hallucination mitigation and expertise retention. To tackle this challenge, we introduce MergeHEIR, a post-merge adaptation framework designed to reduce this merging tax while preserving expertise inherited from initial experts. Using small expert-task calibration sets, MergeHEIR constructs layer-wise null-space projectors via SVD from task-specific activations collected from the merged checkpoint, and periodically projects the accumulated post-merge displacement onto the resulting null spaces to preserve inherited expertise. Theoretically, we establish minimum-distortion and maximum-dimensionality guarantees, characterize the threshold-controlled adaptation-retention trade-off, and extend perturbation guarantees beyond finite calibration data. Across 24 paired comparisons spanning three MLLM configurations and 8 model-merging methods, MergeHEIR consistently mitigates
hallucination while largely preserving inherited expertise, demonstrating a more favorable
hallucination-retention trade-off.
Read the full abstract ▾
All authors Jinyu Li, Hao Fang, Zhiming Zhang, Jiawei Kong, Bin Chen, Shu-Tao Xia
Artificial Intelligence
Carnator: Fast Text-to-Video Generation with Generation-Native Compatibility-Guided Cross-Request Reuse
Xingkun Yin, Xuebin Tang, Mingkun Xu et al. (+1 more)
Video diffusion transformers produce high-quality
videos, yet iterative denoising incurs substantial inference latency, limiting interactive and large-scale serving. Most existing acceleration methods focus on individual requests, thereby restricting efficiency gains to redundancy within a single generation trajectory. Recent cross-request reuse offers an additional source of savings, but existing approaches often infer reusability from coarse semantic similarity. This conflates semantic relatedness with generation-level computational compatibility, so aggressive reuse may accept incompatible historical computation while conservative reuse leaves substantial acceleration unrealized. We present Carnator, a cross-request acceleration framework that addresses this challenge by extracting and using generation-native compatibility evidence directly from the model's evolving internal states. Specifically, Carnator performs a lightweight early probe to construct an Early Signature from internal
diffusion states, assessing reuse validity through risk-aware compatibility decisions. The same evidence characterizes reuse scope by localizing target-specific computation and guiding joint reuse of historical latent trajectories and sparse
attention connectivity. Across three text-to-
video backbones, Carnator consistently achieves higher cache-hit end-to-end acceleration than the evaluated cross-request baselines despite more selective cache acceptance, reaching up to 2.17$\times$ speedup while maintaining competitive generation quality.
Read the full abstract ▾
All authors Xingkun Yin, Xuebin Tang, Mingkun Xu, Hongyang Du
Artificial Intelligence · Cross-listed from cs.RO
RE-0: Verified Recursive Improvement of Embodied Code-as-Policy Agents through Local On-Policy Distillation
Jiawei Zhang, Xiangrong Zhang, Rui Song et al. (+3 more)
Code-as-Policy
agents accomplish long-horizon
embodied tasks by generating and executing code, yet continually improving them with teachers that are stronger but not globally reliable remains a key challenge. Existing
distillation methods typically treat the teacher's complete behavior as the supervision target and thus misassign training credit on states where the teacher fails. We propose RE-0, a recursively verified policy improvement framework: rather than assuming that the teacher globally outperforms the student, RE-0 requests local corrections from the teacher on the student's own failure histories and checks in the environment whether each correction is genuinely beneficial; verified corrections yield immediate improvement. Building on this, we propose RE-OPD, which turns verified interventions into supervision for on-policy
distillation. Only counterfactually verified teacher interventions provide distribution-level supervision, weighted by their measured local benefit, and the improvement they induce is projected back into the standalone student, so both where supervision is applied and how much credit the teacher receives co-evolve with the student policy. We further prove that the student's per-round gain is lower-bounded by its verified intervention gain up to verification and projection error terms. Experiments across multiple Code-as-Policy
embodied tasks show that RE-0 improves both teacher-assisted execution and the standalone student, and generalizes to novel
robots and scenes.
Read the full abstract ▾
All authors Jiawei Zhang, Xiangrong Zhang, Rui Song, Huanbin Zhou, Chengye Song, Hongzhou Wang
Artificial Intelligence · Cross-listed from cs.CL
Automatic Speech Recognition for the Basaà Language: A Low-Resource Approach
Sophie Gertrude Ngo Mock, Charles Moudina Varmantchaonala, Paul Dayang et al. (+2 more)
The rapid advancement of Artificial Intelligence (AI) and Natural
Language Processing (NLP) has revolutionized the way humans interact with machines. Among the most impactful developments is Automatic
Speech Recognition (ASR), which enables computers to convert spoken
language into text. Systems such as those built on deep neural networks,
transformer architectures, and self-supervised learning have achieved near-human performance for well-resourced
languages such as English and French. Yet, these advances have disproportionately benefited a small fraction of the world's
languages.
Read the full abstract ▾
All authors Sophie Gertrude Ngo Mock, Charles Moudina Varmantchaonala, Paul Dayang, Jean Michel Nlong II, Christopher Gies
Artificial Intelligence
Opening LLM Judges: Recovering Preference Signals Beyond the Final Verdict
Sourabrata Mukherjee and Sunayana Sitaram
LLM judges are widely used to evaluate model outputs, but their verdicts can be unreliable: a judge may favor the worse answer for its position, length, or other surface features. When a judge is wrong, is the information needed to judge correctly absent from the model, or present in its internal representations but not reflected in the output? We study this across 64 open-weight evaluators and 14
datasets, including causal interventions on 41 judges (editing activations mid-run to see whether the verdict changes). On LLMBar, built so the superficially better answer is the worse one, the verdicts of 50 judges agree with human labels only 0.456 of the time, even after averaging both answer orders. Yet a small probe on the same judges' activations, with no weight updates, reaches 0.846, and 0.686 once surface features such as length and position are residualized out (0.507 with shuffled labels). The gap holds across eight
benchmarks and model families, but is not universal: a score of how well surface features alone predict the human label, computed before any probe is trained, predicts the size of the gain (Spearman rho = 0.90). On rubric tasks that score one answer at a time, leaving no surface cue to exploit, reading the internals gives no advantage. The interventions also show that editing activations mid-network already changes the verdict, before it can be read off directly, and locate the pathways carrying position and length bias. At the same label budget, the recovered signal lets a judge flag cases where it is likely wrong and yields better labels for preference learning. A wrong verdict, then, does not mean the judge lacks the information, and a simple diagnostic shows when it is worth recovering.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.CL
Shared Worlds, Private Minds: Structured Memory for Long-Form Writing as World Creation
Qiuyu Tian, Xiaowen Gu, Hang Su et al. (+9 more)
LLM
agents that write long-form fiction need an explicit
memory of the evolving storyworld to keep new events consistent with established facts. Such
memory must keep heterogeneous narrative information distinct, integrate story developments across granularities, and recover dependencies that a writing request leaves implicit. We present NarraWorld, a structured
memory system for long-form writing that treats
memory construction as world creation. From a shared evidence-grounded graph, NarraWorld derives four connected views: world facts, per-character beliefs, open developments, and hypothetical branches (possible-world continuations). Hierarchical aggregation with atomic closure consolidates events into scenes, plotlines, and plots, keeping each higher-level node traceable to its constituent source spans. For
retrieval, planned reconstruction infers a query's dependencies from the current narrative situation and a preview of
memory, then assembles the relevant records within a
token budget. Across three writing
benchmarks, NarraWorld achieves the strongest aggregate results. Its
memory also transfers to situated role-playing and largely preserves recall on a general-purpose long-term
memory benchmark, paving the way for
agents that sustain coherent storyworlds across diverse narrative tasks.
Read the full abstract ▾
All authors Qiuyu Tian, Xiaowen Gu, Hang Su, Jianghan Chao, Haojie Yin, Fan Guo, Xin Zhang, Jinjing Shen, Ewing Luo, Youyong Kong, Yingce Xia, Zequn Liu