Artificial Intelligence
VETR: Execution-Grounded Self-Improving Agent for Symbolic Graphics Reasoning
Yingbo Ma, Kuangshi Ai, Gordon Dai et al. (+3 more)
Visual
agents for symbolic graphics
reasoning depend on executable world representations, yet continued improvement remains bottlenecked by ambiguous supervision from final renders. Recent agentic systems refine scenes through repeated interaction, but they lack mechanisms to recursively extract actionable evidence from execution traces, overlooking which proposed symbolic edits survived, disappeared, or interfered during program execution. To address this, we propose VETR (Verified Execution Trajectories for Refinement), a framework that enables agentic recursive self-improvement over executable world representations for symbolic graphics
reasoning. Within each task, VETR retains aligned program/render records with bounded runtime observations and introduces an execution refiner that compiles observed writes and state changes into constrained alternative programs. The refiner validates their endpoints and preservation conditions against captured runtime state; the ordinary visual verifier then assesses goal relevance. Repair construction makes no additional
language-model calls, but can require additional executions. Evaluations on BlenderGym (280 tasks) and BlenderBench (27 released tasks) with GPT-5.6-luna and Qwen3.8-27B show that VETR consistently outperforms existing refinement strategies, including BlenderAlchemy and VIGA. Controlled ablations evaluate the contributions of execution refinement and role-asymmetric evidence organization under fixed generator-interaction caps.
Read the full abstract ▾
All authors Yingbo Ma, Kuangshi Ai, Gordon Dai, Chenglong Yu, Chaoli Wang, Hua Shen
Artificial Intelligence
When Faster Decisions Do Not Make Better Agents: An Empirical Study of Jev in Agent Workflows
Yuchen Ling, Shengcheng Yu, Zhenyu Chen et al. (+1 more)
AI
agents depend on repeated decisions to accomplish tasks, with large
language models (LLMs) commonly serving as their core decision makers. The recent emergence of Jev offers an opportunity to make these decisions faster and at lower cost. However, it remains unclear whether such local savings can be translated into efficient and reliable task completion. In this paper, we investigate this efficiency transfer across five
agents in two representative application settings, embodied tasks in ALFWorld and mobile app use in AndroidWorld. Across 4,500 task executions, we jointly evaluate task effectiveness, decision-level and task-level resource use, and execution behavior. Experimental results show that Jev integration substantially reduces decision cost and latency across all five
agents, with task-level gains varying by workflow. Task success remains close to baseline levels on ALFWorld and declines more substantially in two mobile
agents. Auxiliary
reasoning and additional interactions absorb decision savings, while failed recovery and premature completion limit successful delivery. Mean cost per successful outcome increases in one
agent even as spending per attempt falls. These findings highlight the importance of preserving progress, supporting recovery, and verifying completion when integrating efficient decision models into
agent workflows. Their practical value should be assessed through successful task delivery, recovery behavior, and resources per delivered outcome.
Read the full abstract ▾
All authors Yuchen Ling, Shengcheng Yu, Zhenyu Chen, Chunrong Fang
Artificial Intelligence
SWEAR: Section-Wise Efficient Audio Reasoning via Self-Rated Difficulty
Yixuan Li, Minghao Lv, Xuenan Xu et al. (+2 more)
Recent work on large audio-
language models (LALMs) explores adaptive
reasoning to balance accuracy and efficiency. However, existing approaches often rely on coarse-grained difficulty estimation, overlooking the diverse factors that determine
reasoning effort. We address this limitation by deriving five-dimensional difficulty scores and compressed
reasoning traces from complementary offline assessments. Based on this characterization, we propose SWEAR, a difficulty-aware
reasoning framework that learns difficulty signals during supervised fine-tuning and uses reinforcement learning to adapt
reasoning allocation across
reasoning stages. Experiments on MMAU-test-mini and MMAR show that SWEAR improves accuracy while substantially reducing generated
tokens, shortening
reasoning for easy instances without sacrificing performance on difficult ones.
Read the full abstract ▾
All authors Yixuan Li, Minghao Lv, Xuenan Xu, Zihao Zheng, Mengyue Wu
Artificial Intelligence
Attention Sees, MLP Refuses: Tracing the Attention-to-MLP Safety Pathway in LLMs
Zixuan Weng, Haoqi He, Weifei Jin et al. (+2 more)
Safety-aligned large
language models (LLMs) remain vulnerable to jailbreaks and can over-refuse benign requests. Understanding these failures requires distinguishing harmfulness recognition from the computation that writes refusal. We reframe this as a pathway question and make it measurable on residual-stream updates in decoder-only
Transformers, separating the harmfulness information a module update exposes from the change it produces in readable refusal evidence. Across 18 Base-Instruct pairs spanning Qwen, Llama, and Gemma, harmfulness is decoded more readily from
attention updates, particularly in early layers, whereas MLP updates contribute more strongly to harm-selective refusal writing in aligned models. Post-training strengthens MLP refusal writing in all 18 pairs without systematically improving harmfulness decodability, while positive MLP writing becomes more concentrated in deeper layers after post-training. Across six
reasoning-enabled models, harmfulness remains decodable during
reasoning while strong refusal writing concentrates at answer onset, a pattern that persists with
reasoning disabled. Causal interventions further support this pathway: resetting downstream MLP updates to their clean values substantially reduces the refusal shift caused by upstream
attention interventions. Building on these findings, we introduce Sentinel, which combines an
attention-derived perception score P with an MLP-derived decision score D. Their joint profiles characterize jailbreaks and over-refusals, and conditional routing with both signals achieves a lower aggregate error rate than single-signal routing and evaluated fixed reminders across eight models.
Read the full abstract ▾
All authors Zixuan Weng, Haoqi He, Weifei Jin, Yue Huang, Yang Bai
Artificial Intelligence
LongHarness: Advancing Agentic Reasoning over 100M-Token Corpora via Stateful Evidence Management
Songtao Huang, Zhijie Zhong, Xiaohan He et al. (+8 more)
Corpus-level long-context
reasoning requires
agents to acquire and integrate evidence across document collections that exceed their context windows. Existing
agent harnesses enable incremental access to these corpora, but acquired evidence and evolving conclusions often remain entangled in a growing interaction history. This implicit state management can lead to two coupled failure modes: unbounded exploration, where
retrieval expands the interaction history without resolving decision-relevant information gaps, and evidence-state drift, where previously acquired evidence or subsequent corrections fail to inform the current conclusion. We present LongHarness, a model-agnostic harness that makes the interaction between acquired evidence and evolving conclusions explicit. Its core mechanism, Evidence-Conclusion Coevolution, uses unresolved conclusions to guide exploration toward missing information, while newly acquired evidence strengthens or revises the conclusions that directed the search. LongHarness couples this process with Evidence Lifecycle Management, which keeps accumulated evidence current and traceable throughout the interaction. Together, these mechanisms focus
retrieval on remaining information needs and keep conclusions aligned with current evidence. We evaluate LongHarness on OfficeQA Pro V1 and V2, AA-LCR, and LongBench v2 using GLM-5.3-Flash and Qwen3.8-Flash. Compared with direct
reasoning using the same backbone, LongHarness improves GLM-5.3-Flash exact match on OfficeQA Pro from 15.8% to 60.2% on V1 and from 4.4% to 61.1% on V2. Against state-of-the-art
agent harnesses, it surpasses Claude Code by 12.0 percentage points on V1 and Codex by 18.9 points on V2. With Qwen3.8-Flash, LongHarness reaches 61.7% on V1, outperforming Claude Code by 8.3 points and Codex by 12.0 points.
Read the full abstract ▾
All authors Songtao Huang, Zhijie Zhong, Xiaohan He, Shuaiyu Zhang, Zhiyin Yu, Li Kang, Heng Zhou, Yutao Fan, Xiangchao Yan, Bo Zhang, Shiyang Feng
Artificial Intelligence · Cross-listed from cs.CV
Evolving Creatives with Clicks: CTR-Driven Advertising image Editing via Attribute Disentanglement and Preference Alignment
Xu Zhang, Haohan Wang, Zheng Zhang et al. (+9 more)
The optimization objective for e-commerce advertising creatives has shifted from aesthetic quality to online business metrics like click-through rate (CTR). Recent methods generate appealing ad images from product images by constructing CTR-based reward signals and refining
prompt generation via Reinforcement Learning. Such paradigms equally penalize all attributes within an underperforming image, even though some are actually beneficial to clicks. In this work, we propose Evolving Creatives with Clicks (ECC), which reformulates creative optimization as fine-grained editing rather than generation from scratch. ECC selectively refines attributes that hurt CTR while preserving those that contribute positively. Specifically, faced with a rich attribute space across massive images, ECC first identifies 28 attributes significantly correlated with CTR via ANOVA and Random Forest regression. Based on attribute contrasts between images with disparate CTRs, we instruction-tune a
prompt model to generate editing suggestions, then refine it with real click preferences since reference-derived suggestions are not uniformly beneficial. Furthermore, we introduce Consistent Preference Optimization (CPO) as ECC's preference optimization strategy, which enforces consistency across semantically equivalent shuffled suggestions to mitigate DPO overfitting and improve preference generalization. Extensive experiments demonstrate that ECC significantly improves the CTR of ad images, outperforming state-of-the-art models by a substantial margin. The complete data pipeline, including attribute extraction and analysis, the resulting
dataset, and source code will be released upon acceptance.
Read the full abstract ▾
All authors Xu Zhang, Haohan Wang, Zheng Zhang, Yingjie Xia, Wei Feng, Honghe Zhu, Jingjing Lv, Changping Peng, Xuheng Qin, Zehua Zhang, Zheng Luo, Qicheng Li
Artificial Intelligence
Travel Back To Move Forward: Retrospective Exploration and Evolving for Long Horizon Agents
Ziang Ye, Wentao Shi, Lei Chen et al. (+7 more)
Large
language model
agents increasingly operate in long-horizon environments where early mistakes can trigger cascading failures and trap
agents in dead-end trajectories.
Agents must therefore be able to proactively recover from such states and turn their exploratory experience into guidance for subsequent decisions. We introduce ReTRAIL, a framework that enables
agents to rewind the environment to earlier states while preserving valid progress and carrying forward knowledge learned from abandoned exploration. The environment state moves backward while the
agent's knowledge continues to move forward. ReTRAIL trains a shared policy with two complementary signals: State-Anchored GRPO compares alternative continuations from revisited states to improve execution, while a branch-advancement reward guides reflection through the progress of subsequent exploration. Experiments across four interactive
benchmark MineSweeper, Sokoban, ALFWorld and ScienceWorld show that simply adding a rewind primitive provides limited benefit and can even degrade performance. In contrast, ReTRAIL enables
agents to rewind from unproductive states and use insights from abandoned branches to guide subsequent exploration. These results demonstrate the value of coupling environment rewinding with experience reuse within an ongoing task for long-horizon task solving. The code is available at https://github.com/whi497/ReTRAIL.
Read the full abstract ▾
All authors Ziang Ye, Wentao Shi, Lei Chen, Zhengzhou Cai, Yu Wang, Zhiyuan Yao, Zihan Wang, Xierui Song, Fuli Feng, Yixin Cao
Artificial Intelligence · Cross-listed from cs.CL
CoBe: A Benchmark For Conversational Counterfactual Text Editing
Arvind Raghavan, Inwoo Hwang, Shreyas Havaldar et al. (+6 more)
Counterfactual text editing is the ability to revise a given text in light of a hypothetical change while preserving facts not causally affected by that change. When asked what would have happened if a doctor had not recommended surgery, the model output should retain the patient's medical history and modify only the consequences of the decision. This requires causal understanding: a model must determine, without being explicitly instructed to, what events are affected by the hypothetical change and what should remain fixed. This skill is central to human
reasoning, especially in safety-critical domains like healthcare or law. By contrast, the task of associational text generation merely generates text statistically associated with the altered condition. In the medical example, an associational response might make the patient's symptoms less severe because milder cases correlate with no surgery. Such responses produce a plausible scenario, but not the counterfactual version of the given one. Prior counterfactual-text
datasets do not clearly separate these modes of
reasoning. We introduce the Counterfactual
Benchmark (CoBe), a testbed for models' zero-shot ability to counterfactually edit a given textual scenario when prompted to do so in a conversationally natural way. CoBe contains 2000 questions spanning healthcare, science and engineering, business, and everyday scenarios. We test a suite of frontier and smaller models, with frontier systems achieving an average score of $58.05\%$. Our findings indicate that models frequently default to associational generation rather than preserving the causal structure of the original scenario. Moreover, performance does not scale consistently with either model size or output
token length, suggesting that the failure may reflect model training objectives rather than insufficient model capacity.
Read the full abstract ▾
All authors Arvind Raghavan, Inwoo Hwang, Shreyas Havaldar, Kai-Zhan Lee, Aurghya Maiti, Yushu Pan, Jeffrey Wu, Mingxuan Li, Elias Bareinboim
Artificial Intelligence · Cross-listed from cs.CV
GeoGR: Spatial-Semantic Token Compression and Group Routing for Efficient 3D VLMs
Yongxin Zheng, Hao Wu, Fang Li et al. (+8 more)
Multi-view representations enable pretrained image and
video vision-
language models to understand 3D scenes, but their long visual
token sequences make inference costly. Redundancy arises both from repeated observations of the same scene regions across views and from processing all visual
tokens at every LLM layer, regardless of their relevance to the current computation. We propose GeoGR, a two-stage framework that jointly compresses scene representations and visual computation. At the projector level, GeoSemZip consolidates cross-view observations into 3D voxel
tokens and combines semantic saliency with spatial coverage to construct a compact scene representation. Unselected voxel features are aggregated into contextual anchors to retain scene information that would otherwise be discarded. At the LLM level, we identify an effective visual computation window and observe that query-conditioned
token relevance remains locally stable but changes over longer layer intervals. Guided by these findings, GroupRoute combines late entry and early exit with recoverable group-wise skipping. Low-relevance
tokens bypass complete LLM layers while retaining their hidden states and can rejoin the active group at subsequent anchor layers selected under a computation budget. For compression-aware post-training, we construct a compact
dataset of 67K samples by retaining 30% of the original training pool. With this protocol, GeoGR retains 95.2% and 96.8% of the dense post-trained models' performance on LLaVA-OneVision-7B and
Video-3D LLM-7B, respectively, while accelerating LLM prefill by $6.8\times$ and $6.1\times$. These results demonstrate the consistent effectiveness of GeoGR across 3D VLMs with and without explicit 3D spatial encoding, highlighting a favorable performance-efficiency trade-off. The code is available at https://github.com/EIT-NLP/GeoGR.
Read the full abstract ▾
All authors Yongxin Zheng, Hao Wu, Fang Li, Hu Fu, Peiran Yin, Xudong Wang, Haozhe Hu, Xinghao Chen, Yunpu Ma, Wei Zhang, Xiaoyu Shen
Artificial Intelligence · Cross-listed from cs.SE
Don't Guess, Just Ask: Evaluating Interactive Requirement Clarification for Repository-Level Software Development
Mingyu Chen, Zhiqing Zhong, Manyi Wang et al. (+3 more)
As LLMs become increasingly capable of implementing software, a critical bottleneck emerges: whether they truly understand what users want. Users often begin with vague intent, leaving important requirements to be clarified through interaction. In repository-level development, misunderstandings can propagate and cause costly rework. However, existing
benchmarks typically evaluate implementation from complete specifications or underspecified requests, overlooking pre-coding requirement clarification. Prior clarification studies mainly focus on function-level tasks. It therefore remains unclear how well LLMs can transform vague intent into clear requirements through interaction and how this capability affects subsequent code generation. We introduce ReqClarifyBench, an interactive requirement clarification
benchmark for repository-level software development. It contains 341 tasks spanning game, miniapp, and web development, each with a vague user intent, a human-validated requirement document, and code evaluation artifacts. We employ LLM-based user simulators with diverse personas derived from real-world conversations. We evaluate six LLMs in terms of clarification effectiveness and generated requirement document quality. Results show that stronger clarification effectiveness does not necessarily translate into higher document quality, and performance varies substantially across user personas. We also find that coding-
agent harnesses do not further improve requirement document quality compared with their base LLMs. These findings offer insights for building LLM-based end-to-end software development systems.
Read the full abstract ▾
All authors Mingyu Chen, Zhiqing Zhong, Manyi Wang, Linxi Liang, Rouyu Chen, Pinjia He
Artificial Intelligence
EmbodiedCLBench: Evaluating Continual Learning for Self-Evolving Embodied Agents
Jie Huang, Yanan Chen, Ruixun Liu et al. (+6 more)
Continual learning requires models to accumulate past experience and transfer it to new tasks. Despite substantial research on continual learning, existing work remains largely confined to non-embodied settings, without extending to embodied environments. To address this gap, we introduce EmbodiedCLBench, the first
benchmark in embodied simulation for evaluating the continual learning capability of self-evolving
agents. With compositional generalization as the core evaluation principle, our
benchmark evaluates whether
agents can first learn from basic tasks and then recombine the acquired skills to solve complex, unseen advanced tasks. Based on EmbodiedCLBench, we conduct extensive experiments across multiple
agent harnesses and learning methods, revealing consistent limitations and distinctive behaviors. Our results show that while advanced tasks can be hardly solved under zero-shot setting, the experience from basic tasks enables consistent improvement. Such benefit from the experience remains stable even if the task complexity increases. However, this improvement does not scale accordingly with the number of learning samples, and current
agents cannot extract meaningful information from additional experience. Collectively, our findings highlight continual learning as a fundamental yet underexplored capability, and our
benchmark offers valuable resources for designing effective self-evolving
agents in embodied environments. Our code and
dataset are available at https://pku-value-lab.github.io/EmbodiedCLBench-homepage.
Read the full abstract ▾
All authors Jie Huang, Yanan Chen, Ruixun Liu, Kaichen He, Xinyi He, Rui Huang, Zilong Zheng, Zhenliang Zhang, Yiwu Zhong
Artificial Intelligence · Cross-listed from cs.LG
Generalized Proximal Policy Optimization: Unifying LLM RL Algorithms from the Perspective of Divergence
Zhicheng Cai, Xinyuan Guo, Hanlin Wu et al. (+5 more)
Reinforcement learning has become the foundational paradigm for enhancing the
reasoning capabilities of LLMs. Yet the dominant approach, Proximal Policy Optimization (PPO), relies on a geometrically degenerate clipping proxy that cannot adapt its resistance to policy shifts, leading to premature clipping and degraded exploration. Subsequent variants alleviate these issues through fragmented and heuristic modifications. We propose Generalized Proximal Policy Optimization (GPPO), a principled and unified framework for understanding and designing LLM RL
algorithms through the lens of divergence. We establish a generalized $f$-divergence trust-region theory with a monotonic performance improvement guarantee, and show that different $f$-divergences induce distinct geometries and policy update dynamics. We further introduce novel
estimator topologies that provide more faithful divergence estimates than the conventional single-sample Monte Carlo estimation. Under this framework, PPO and recent variants such as DAPO, DPPO, and RIPO emerge as special cases, establishing a principled two-dimensional design space of {Divergence $\times$
Estimator}. Extensive experiments on eight competition-level math and coding
benchmarks show that both dimensions are critical. The best-performing GPPO instantiation surpasses standard GRPO by up to 50% and 14% on math and coding, delivering SOTA performance.
Read the full abstract ▾
All authors Zhicheng Cai, Xinyuan Guo, Hanlin Wu, Weinan Dai, Yuyan Zhou, Wei-Ying Ma, Ya-Qin Zhang, Hao Zhou
Artificial Intelligence
Provenance-Aware Belief Updating in Human-AI Feedback Loops
Tianyi Huang, Eric Liu, Jay Chong et al. (+2 more)
In repeated human-AI interaction, a single observation can be quoted, summarized, and returned to a model as multiple reports, making repetition look like independent corroboration. We study whether
language models can use supplied report lineage to update according to evidence provenance rather than report count. We introduce Latent Quotient Evidence Training (LQET), a relational objective that combines sparse
Bayesian supervision with constraints that encourage stable predictions when reports add no new evidence and consistent updates to the same new evidence across relay histories. On a controlled binary
benchmark, LQET reduces new-evidence update RMSE by 41-44% and log-odds drift after eight relay steps by 89-91% relative to task-only training across Qwen3-4B and Phi-4-mini, with three paired training seeds per backbone. The gains extend to a closed-loop simulation in which model confidence controls how many relayed reports re-enter the model's context in later rounds. By grounding belief updates in supplied provenance, LQET offers a route toward dynamic human-AI systems in which influence follows new evidence, not the number of times the same evidence returns.
Read the full abstract ▾
All authors Tianyi Huang, Eric Liu, Jay Chong, Emily Yin, Wanqi Zhao
Artificial Intelligence
Scaling Empowered Agents
Catherine Ji, Vivek Myers, Joshua Lin et al. (+3 more)
Empowerment measures the capacity of an
agent to actively control its environment. In this work, we explain the connection between skill learning and empowerment. While skill learning is not empowerment, we introduce a simple scalable recipe for empowerment maximization: empowerment corresponds to the number of executable skills from a state, so maximizing empowerment corresponds to reaching states that are good starting points for many skills. Our recipe applies to arbitrary underlying skill learning methods. Whereas most prior work on empowerment looks at tabular settings, this recipe enables us to study empowerment at a scale where we can realize theoretically-predicted but empirically-elusive properties of empowerment (e.g., gathering resources, environment modification, survival). Experiments show that empowerment maximization unlocks interesting emergent properties, such as building a house in Crafter. Website and code at \href{https://empowerment-geometry.github.io/scaling-empowerment/}{this https url}.
Read the full abstract ▾
All authors Catherine Ji, Vivek Myers, Joshua Lin, Alexandra Wingate, Sergey Levine, Benjamin Eysenbach
Artificial Intelligence
Hindsight Structured Advice Distillation: Steering Executor Agents with Advisors Trained on Their Failures and Task Characteristics
O. Ozan Koyluoglu
Language-model
agents are used for specialized work such as financial research, healthcare and law. Such
agents are often capable enough for a task but still miss its details. In finance, for example, these details include which filing reports a figure or how a company discloses a metric. Retraining the
agent to learn these details is expensive, and often impossible when the
agent is a closed or shared system. Writing the missing know-how into the
agent's
prompt makes the
prompt long and identical for every input, although different inputs need different guidance. We propose to learn to generate this guidance instead. A small advisor model reads each input and writes structured advice, which is appended to the
prompt of a frozen executor. The advisor is trained with hindsight, following work on
distillation with privileged information. On a training pool with graded answers, we analyse the executor's failures, run it under several strategies to see which one worked on each input, and let a teacher that sees the grades and the traces write the advice that would have helped. The advisor then learns, by supervised fine-tuning, to write that advice from the input alone. Because the student is a separate small model rather than the executor, it is cheap to train and adds little latency. We call this approach hindsight structured advice
distillation (HSAD). On FrontierFinance, a public
benchmark of expert-written financial research queries, an 8B advisor raises the qualification rate of Samaya's proprietary research
agent, which already leads the
benchmark's leaderboard, from 52.9% to 56.6%. This is above the
agent's own high-effort configuration, at 37% lower cost per query. We study the generalization of the proposed approach, and in particular show that HSAD also improves results on WebShop, BIRD text-to-SQL and ExpertLongBench.
Read the full abstract ▾
Artificial Intelligence
Benchmarking Embedding Geometries and Clustering Algorithms for Unsupervised Patent Claim Classification
Sujith Subramanian, P N Bhargav Teja, Susen Sarkar et al. (+1 more)
Prior-art
retrieval systems are structurally organized around classification codes, making cross-code discovery difficult. This paper asks if an unsupervised representation trained on patent text can recover relevant connections that code-restricted search cannot, and if
hyperbolic space helps. We trained a
hyperbolic and a matched Euclidean Variational Autoencoder (VAE) on 45,000 US patents (96 runs, 6 seeds). The
hyperbolic model retrieves genuinely cross-category prior art 52% more often than the Euclidean model (18.76% vs. 12.34% of top-10 results, 3,000 queries). This represents a practical gain for the motivating search problem. The
hyperbolic model agrees less with official classification (NMI 0.2392 vs. 0.3206), which we explain geometrically: the two measures capture different goals, and only one aligns with prior-art search needs. In a study of 10 technology domains (50,000 more patents), we find that a domain's classification breadth–the number of distinct CPC subgroups spanned–predicts approach performance (Pearson r=-0.642, p=0.045, N=10), and that patents per subgroup is an even stronger predictor (r=0.715, p=0.020). This provides a checkable signal for practitioners to determine where the method is most valuable. We report all findings with statistical transparency and release a complete, 91-test automated implementation.
Read the full abstract ▾
All authors Sujith Subramanian, P N Bhargav Teja, Susen Sarkar, Rajasekhara Reddy Duvvuru Muni
Artificial Intelligence · Cross-listed from cs.CL
What Should SFT Learn? Conditional Information Gain for Generalizable Post-Training
Yuxuan Gu, Kequan Niu, Xiaocheng Feng et al. (+3 more)
Supervised fine-tuning (SFT) remains a widely used approach for post-training large
language models, because fixed demonstrations can directly introduce desired
reasoning strategies and behavioral patterns. Yet forcing a model to imitate externally provided trajectories may also interfere with behaviors already supported by pretraining, limiting generalization beyond the training distribution. Recent dynamic SFT methods mitigate this issue by reweighting
token-level supervision according to model-side signals such as predictive probability, prior support, or distributional compatibility. However, model compatibility is not equivalent to learning value: a
token that is weakly supported by the current model may either be an undesirable mismatch or precisely the task-specific decision that the demonstration is intended to teach. We propose CIG-SFT, an information-theoretic framework that estimates the task-conditioned information carried by each demonstrated
token. Under teacher forcing, we compare the
token likelihood from a frozen reference model with and without access to the task input while keeping the gold solution prefix fixed. The resulting conditional information gain quantifies how much the task itself reduces uncertainty about the demonstrated decision. We then convert this information into a KL-regularized projection of the current policy and derive an equivalent scalar-weighted SFT objective whose logit gradient exactly matches that of the projected target. Across mathematical
reasoning and code-generation
benchmarks, CIG-SFT consistently outperforms standard SFT and recent adaptive fine-tuning baselines, with particularly strong improvements in generalization to challenging evaluation distributions, while requiring neither additional training data nor online rollouts.
Read the full abstract ▾
All authors Yuxuan Gu, Kequan Niu, Xiaocheng Feng, Yun Li, Yiqiao Yang, Bing Qin
Artificial Intelligence
Work2Skill: Capturing Human GUI Experience for Reusable Agent Skill Induction
Yuqi Huang, Hui Li, Yifei Cao et al. (+9 more)
Computer-use
agents can perform increasingly complex GUI tasks, yet unfamiliar software and workflows requiring human domain knowledge remain challenging. Human demonstrations convey procedural knowledge through the same GUI interaction space, but reuse requires interpreting actions in context and separating workflow rules from task-specific details. We thus introduce Work2Skill, a framework that captures human GUI experience for reusable
agent skill induction. WorkRecorder provides the framework's infrastructure for collecting this experience, aligning screen recordings, mouse and keyboard events, and available accessibility-tree information. Using WorkRecorder, we construct WorkDemo-350, the first synchronized
multimodal dataset of human desktop GUI demonstrations for reusable
agent skill induction and downstream evaluation. Its 350 demonstrations are screened for completeness and task success. We then propose an agentic skill-induction method that inspects evidence on demand to produce reusable skills capturing workflow rules, step dependencies, and completion checks. When the same four execution models are equipped with skills induced by each method under shared execution settings, Work2Skill achieves the highest scores and success rates. Mean score improvements over the corresponding No Skill baseline are 4.67–10.54 points on MyPCBench Bounded Action and 3.11–7.50 points on a 286-task subset of OSWorld-Verified. Our code and data are available at https://github.com/china-qijizhifeng/Work2Skill.
Read the full abstract ▾
All authors Yuqi Huang, Hui Li, Yifei Cao, Zikun Cai, Tianci Xie, Yangfan Gao, Pinxue Guo, Chaoyun Zhang, Bingxuan Li, Rui Zheng, Hang Yan, Ning Liao
Artificial Intelligence · Cross-listed from cs.CV
The Evidence Anchoring Boundary: Where Visual Interventions Matter and When They Help in Multimodal LLMs
Qiji Zhou, Junyi Hu, Peiyu Wang et al. (+1 more)
Visual interventions provide a training-free way to reduce
hallucinations in
multimodal large
language models. However, an intervention may have little leverage at some layers and may damage a correct answer even when it changes the output. We investigate where visual interventions retain leverage and when they improve predictions. First, we apply the same visual-residual edit at successive decoder layers, matching its RMS strength across depth. Across six checkpoints from three model families, we observe a persistent drop in intervention effect and define its operational onset for this edit as the Evidence Anchoring Boundary (EAB). With boundaries fixed on calibration queries, held-out effects on disjoint evaluation queries before EAB are $13.78\times$–$26.10\times$ larger than those after it. Comparison with intermediate readouts shows that the edit can lose leverage before answer preferences stabilize. Second, we compare counterfactual subtraction across presence, attribute, and relation questions. Subtraction can correct a misleading preference, leave a decision unchanged, or remove a preference supporting the correct answer. Based on these observations, we develop calibrated selective routing, which preserves the original prediction unless calibration supports an intervention for the question type. It improves full-set attribute accuracy by $1.56$–$1.81$ percentage points across three checkpoints, while universal subtraction is neutral or harmful.
Read the full abstract ▾
All authors Qiji Zhou, Junyi Hu, Peiyu Wang, Yue Zhang
Artificial Intelligence
Beyond Semantic IDs: Finite-Beam Reachability and Continuous-Query Generation in Generative Recommendation
Yu Sun, Renkai Xiang, Zuodong Xiang et al. (+6 more)
Generative recommendation (GR) of social media platforms commonly retrieves items by autoregressively decoding discrete Semantic IDs (SIDs) with finite-width beam search. This work examines how this output interface constrains whether held-out targets enter finite returned lists, while distinguishing catalog-wide scoreability from empirical top-K support. We introduce two continuous-query paradigms: CQG-Single generates one
retrieval query in a single pass, and CQG-AR autoregressively generates continuous residual updates, retaining step-wise decoding while removing the SID item-
token vocabulary, code lookup, and discrete beam expansion. These designs thereby eliminate the SID-specific reachability constraint induced by finite-beam decoding and avoid the technical challenge of aligning a discrete SID vocabulary with the pretrained backbone's vector space. Both methods make every encoded item directly scoreable through dense or approximate-nearest-neighbor
retrieval. Under a unified protocol with valid-prefix constrained decoding, CQG-Single exceeds the TIGER-style baseline in all seven runs while using 55.3% as many trainable parameters. CQG-AR remains comparable on ML-1M and is numerically stronger on Beauty and Sports
datasets. A controlled depth ablation further shows that additional continuous refinement does not monotonically improve shallow accuracy. These findings motivate decoupling output representation, autoregressive modeling, and
retrieval budget when evaluating generative recommendation systems.
Read the full abstract ▾
All authors Yu Sun, Renkai Xiang, Zuodong Xiang, Junwei Zhou, Yike Zhang, Jianxin Zhang, Zhihao Zhu, Xiaoyu Cao, Hailu Xu
Artificial Intelligence
SmartSalesBench: A Resistance-Aware Multi-Turn Benchmark for Evaluating LLM Sales Agents
P N Bhargav Teja, Velpula Lokesh Reddy, Madhu Basumata et al. (+2 more)
We present SmartSalesBench, the first multi-turn conversational
benchmark explicitly designed to evaluate large
language model (LLM) sales
agents against psychologically grounded, resistant buyer personas. Existing task-oriented dialogue
benchmarks assume cooperative users and measure task completion; SmartSalesBench inverts this assumption, placing
agents against personas programmed to resist and challenge across conversations of up to twenty turns. The
benchmark includes five
agent profiles, twelve buyer personas, fifteen smartphone products across six price tiers, and four controlled experiments with 3 iterations totaling 13,680 sessions. Both sales-
agent and buyer persona generations are grounded through
retrieval-augmented generation (RAG). We introduce nine evaluation metrics grouped into three tiers: primary outcome metrics (CR, BRS, RIR), secondary quality metrics (SMS, PFS, ETI), and exploratory metrics planned for future reporting (ITS, ARI, SPI). Our evaluation combines automated LLM-as-judge scoring with internal human assessment (n = 50, ąppa} = 0.71, Pearson r = 0.81 with BRS).
Read the full abstract ▾
All authors P N Bhargav Teja, Velpula Lokesh Reddy, Madhu Basumata, Susen Sarkar, Rajasekhara Reddy Duvvuru Muni
Artificial Intelligence · Cross-listed from cs.CV
MotionCoach: Constructing Multimodal Guidance for Video Professional Motion Editing
Siyi Qian, Hanqing Yang, Weihuang Lin et al. (+4 more)
Video editing models achieve strong performance in modifying appearance, backgrounds, styles, and motions, empowered by large-scale training data. Beyond general scenarios, professional motions are significant in performance, sports, and martial arts. This paper focuses on professional motion editing, which involves fine-grained motion technique and temporal structure. Given professional motion names but lacking detailed descriptions, existing models often fail to generate professional motions, such as incorrectly lifting the front wheel under a Stoppie instruction. To address this challenge, we introduce MotionCoach, a trained
multimodal agent that uses multi-turn tools and performs self-verification, producing a grounded
prompt and a reference clip as reliable and sufficient
multimodal guidance to inject external professional motion knowledge into
video editing models. To supply training data, we build a data pipeline to construct trajectories. We cold-start the
agent on MotionCoach-SFT-10k and apply Group Relative Policy Optimization (GRPO) on MotionCoach-RL-2k with rewards on the motion guidance
prompt, the motion outcome
video, and the format. For evaluation, we introduce VPMEBench, comprising 314 professional motion editing tasks with fine-grained references and annotations. Extensive experiments show that MotionCoach improves the overall score from 55.1 to 65.6 on Seedance 2.0 and from 51.2 to 64.5 on MiniMax-H3. Learning to construct
multimodal guidance enables precise professional motion editing without retraining
video editing models.
Read the full abstract ▾
All authors Siyi Qian, Hanqing Yang, Weihuang Lin, Zihao Yang, Yichen Lu, Yuanhong Zheng, Sheng Li
Artificial Intelligence · Cross-listed from cs.RO
JoyInside-VLM-Agent: Product-Conditioned Multimodal Policy Learning for Embodied Agents
Gangyi Li, LuYao Wang, Jingfu Sun et al. (+7 more)
Embodied
agents must translate
multimodal observations and user requests into executable actions while adapting to product constraints, changing goals, and execution feedback. We present JoyInside-VLM-
Agent, a ReAct-style
multimodal agent developed through product-conditioned trajectory construction and staged policy learning for robotic platforms and smart hardware. A closed-loop sandbox generates
multimodal interaction trajectories covering goal revisions and failure recovery, which are checked against execution outcomes and semantic criteria. Supervised fine-tuning and multi-teacher on-policy
distillation precede reinforcement learning with DPER (Dual-Source Privileged On-Policy Self-
Distillation with
Token-Level Evidence Routing). DPER routes
token-level evidence between a frozen privileged teacher and a privileged student sharing the rollout policy's parameters, modulating reward-derived
token advantages while preserving their signs. Training rewards combine generative process evaluation, multi-judge rankings, and task-queue consistency. JoyInside-VLM-
Agent achieves a BFCL v4 overall score of 77.44, exceeding the evaluated baselines; results on the other public
benchmarks are mixed. On JoyInside-Devices and JoyInside-Robotics, the evaluated system, which includes domain-specific skill refinement, leads the baselines on all reported metrics. Two real-world quadruped case studies illustrate visually grounded search, replanning after execution failure, and explicit failure reporting.
Read the full abstract ▾
All authors Gangyi Li, LuYao Wang, Jingfu Sun, Jingyi Yuan, Hang Liu, Hao Zong, Yujia Hu, Tingting Gao, Jianyong Wang, Xu Yang
Artificial Intelligence
MathVis: Can AI Reason through Drawing? A Benchmark for Mathematical Figure Generation
Jingjing Wei, Hao Zhao, Xinke Li et al. (+3 more)
This paper introduces MathVis, a
benchmark designed to evaluate LLMs' ability to visualize frontier mathematical problems and perform diagram-assisted
reasoning. The
benchmark consists of 512 tasks, each comprising a mathematical statement, a reference diagram, executable TikZ or SVG code, a Gold Intermediate Representation (IR), atomic rubrics, and error codes. Through rigorous evaluation, GPT-6-Astra, Claude Fable 5, and Gemini 3.1 Pro Preview achieve pass@1 scores of 31.25%, 26.56%, and 22.85%, respectively. We found that although the tasks require visualization, LLMs produce outputs dominated by formal
languages, symbols, and graphics programs, rather than explicit diagram-mediated
reasoning. Furthermore, while the solutions generated by LLMs may contain all the necessary elements, they are often visually confusing to humans. Ultimately, this
benchmark aims to visualize mathematical problems and promote research on enabling LLMs to simplify their
reasoning via visual diagrams.
Read the full abstract ▾
All authors Jingjing Wei, Hao Zhao, Xinke Li, Linxi Yu, Xiaosong Gao, Shanshan Xiao
Artificial Intelligence · Cross-listed from cs.CL
Surfing the Activation Wave: Analyzing the Information and Geometry of LLM Hidden States for Speculative Decoding
Raghavv Goel, Sudhanshu Agrawal, Mukul Gagrani et al. (+2 more)
Decoder hidden states from different layers of large
language models (LLMs) often encode complementary information. However, existing speculative decoding methods typically rely on a small set of target-model layers chosen using fixed, position-based heuristics. Selecting an optimal subset of layers is inherently a combinatorial problem, making exhaustive evaluation infeasible even for moderately deep models. In this work, we present a principled framework for multi-layer hidden-state selection for speculative decoding based on two complementary criteria: informativeness, measured through
matrix-
entropy analysis, and diversity, quantified using centered kernel
alignment (CKA). We investigate layer-wise Shannon
entropy, aggregate
entropy under hidden-state concatenation, and pairwise CKA dynamics across depth, revealing distinct
geometric and informational compartments within decoder representations. Motivated by these findings, we propose
entropy and CKA-based layer selection
algorithms that favor hidden-state sets that are both highly informative and minimally redundant. As part of this framework, we introduce
token-
entropy filtering, which removes hidden states associated with low predictive-
entropy tokens, thereby improving the quality of layer-selection statistics. In EAGLE3-style speculative decoding, our selected sets outperform fixed baseline choices, including the default $[2, N//2, N-3]$ rule, while providing a clearer explanation of why some defaults can remain competitive. Our results provide a practical recipe for multi-layer hidden-state selection for speculative decoding, insights into the geometry and information in LLM decoder states, and a foundation for extending similar analysis to related applications.
Read the full abstract ▾
All authors Raghavv Goel, Sudhanshu Agrawal, Mukul Gagrani, Risheek Garrepalli, Mingu Lee
Artificial Intelligence
Learning More by Training Less: Uncertainty-Guided Progressive Contraction for Efficient Reasoning Post-Training
Suorong Yang, Jiaxin Wang, Jiaming Su et al. (+4 more)
Large
language models have achieved remarkable advances in complex
reasoning. However, scaling
reasoning post-training with ever-larger corpora can substantially increase training cost without proportional performance gains. In this study, we therefore ask: how far can
reasoning supervision be compressed without sacrificing capability? To this end, we propose an uncertainty-guided progressive contraction framework that concentrates supervision and optimization on the
reasoning trajectories most valuable for learning. Specifically, rather than treating a teacher-generated chain-of-thought (CoT) as a single realized path, we construct uncertainty-profiled
reasoning supervision by building Monte Carlo
reasoning trees around intermediate
reasoning states to expose process-level uncertainty. This uncertainty first contracts the original corpus into a compact teacher-informed pool. A second, learner-aware contraction then performs online routing during SFT, concentrating optimization according to the student's evolving learning state. Together, the two stages progressively contract both supervision and optimization, substantially reducing the student training budget. Experiments across diverse
reasoning benchmarks and LLM families show that our method uses as little as 0.22% of the optimization budget of full-data SFT, over 450x less training exposure, while maintaining or improving
reasoning performance. Notably, our framework generalizes well across heterogeneous student families, black-box teachers, and weak-to-strong transfer. This indicates that it is not only a way to accelerate training but can also enhance
reasoning capability.
Read the full abstract ▾
All authors Suorong Yang, Jiaxin Wang, Jiaming Su, Chentao Yue, Furao Shen, Peng Ye, Soujanya Poria
Artificial Intelligence
Collaborative Learning and Reasoning over Model Slices
Yuheng Wang, Runde Yang, Zhuohan Bao et al. (+9 more)
Large
language models (LLMs) excel across diverse tasks, benefiting from knowledge priors, in-context
reasoning, and test-time computation. Despite these strengths, single-model
reasoning remains vulnerable to error propagation, as later steps depend on the model's earlier outputs, leaving
reasoning blind spots difficult to independently reassess. To address this limitation, we develop collaborative learning, encompassing the construction of heterogeneous collaborators and the learning of effective cross-model collaboration. Specifically, model slicing replaces a large model with smaller, heterogeneous models under a comparable parameter budget, enabling prior
reasoning to be reassessed across distinct model parameterizations. Pyramid supervision combines protocol
distillation with step-level supervision to refine
reasoning and interaction decisions throughout collaboration. We evaluate our approach across four task categories: multi-hop question answering, mathematical
reasoning, code generation, and open-ended constrained generation. The trained heterogeneous models achieve comparable or better task performance than the original large model and established multi-
agent baselines, including reinforcement-learning-based methods, while also achieving higher inference efficiency than the multi-
agent baselines and exhibiting differentiated collaboration patterns across tasks. These findings show that learned collaboration turns model heterogeneity into complementary
reasoning capabilities.
Read the full abstract ▾
All authors Yuheng Wang, Runde Yang, Zhuohan Bao, Qiran Zhang, Huatao Li, Yiyang Zheng, Lin Wu, Jingru Fan, Shu Yao, Yihan Li, Zhi Li, Chen Qian
Artificial Intelligence
Teaching neural decoders with language
Xuan Wen, Yonghao Song, Yuhong Zhang et al. (+3 more)
Brain-computer interfaces decode neural activity, yet class-label supervision (CLS) leaves relationships among classes implicit. We propose
Language as Supervision (LaS), which aligns neural features with natural-
language class descriptions through learned projections and predicts classes using candidate text prototypes. We evaluate seven encoder architectures with independently trained models for visual, motor, and
speech decoding from scalp and intracranial recordings. LaS improves
speech-decoding accuracy by up to 17.15~pp over matched CLS baselines. Its representations support candidate-set reuse and label-supervised subject and session adaptation, while LaS enables conditional unseen-frequency decoding in steady-state visual evoked potential (SSVEP) responses. Correct class-description correspondence supports this generalization, and dual-sided attribution visualization shows neural and textual contributions to the same class-matching score. These findings suggest that text descriptions can serve as general-purpose references for learning and interpretation in neural decoding.
Read the full abstract ▾
All authors Xuan Wen, Yonghao Song, Yuhong Zhang, Guoliang Li, Tzyy-Ping Jung, Xiaorong Gao
Artificial Intelligence
AI-Mediation in Retail Markets: Revenue Effects and Assortment Optimization
Omar El Housni and Marouane Ibn Brahim
AI shopping assistants increasingly influence which products consumers consider. We study retail assortment planning under AI mediation, where a retailer chooses a catalog, an AI intermediary forms a customer-facing shortlist, and the customer chooses from that shortlist according to a multinomial logit model. We consider both deterministic mediation, in which the AI follows a fixed segment-specific ranking, and
stochastic mediation based on the Plackett-Luce model. Empirically, we construct a
dataset of product rankings from leading AI models, estimate Plackett-Luce ranking models, and evaluate their predictive performance on held-out assortments. Combining these estimates with synthetic customer preferences, we find that accounting for AI mediation can materially change optimal assortments and revenue. Across the tested capacity configurations, ignoring mediation yields mean revenue losses between 2.52% and 11.37% relative to the PL optimum. A much simpler deterministic plug-in, based only on the ordering of fitted AI utilities, performs substantially better, with mean losses of at most 0.53%. Motivated by this performance, we study assortment optimization under deterministic mediation. We characterize its computational complexity and show that, for any fixed number of customer segments, the problem admits a fully
polynomial-time approximation scheme.
Read the full abstract ▾
Artificial Intelligence
DeFiRank: Federated Pairwise Ranking for Governance in Decentralized Financial Networks
Zahra Batool, Michael Duchesne, Kaiwen Zhang et al. (+1 more)
Existing governance approaches in decentralized financial networks rely on manual review processes conducted by human committees, which introduces subjective bias, creates a scalability bottleneck under high network volume, and fails to capture complex behavioural patterns and evolving network dynamics. To address these limitations, we propose \name, a
federated learning-based pairwise ranking framework in which multiple institutions collaboratively train a ranking model without sharing their private transaction data. This collaborative learning enables the model to capture network-wide behavioural patterns that signify strong network participation and directly contribute to the network effect, including diversity of counterparties, usage of multiple applications, and frequency of transactions. The proposed system ranks providers based on the authenticity of their actions, promoting genuine contributors and demoting those exhibiting suspicious behaviour. We evaluate \name\ on a real-world blockchain transaction
dataset. Results demonstrate that the
federated approach achieves ranking performance modestly below the centralized baseline (NDCG@15: $\approx$0.87 vs.\ $\approx$0.92) while preserving data
privacy and matching its ranking stability (Kendall $\tau$: $\approx$0.70 for both regimes), producing reliable governance decisions, including provider promotion and demotion (648 and 39 unique providers, respectively), without requiring centralized access to sensitive transaction data.
Read the full abstract ▾
All authors Zahra Batool, Michael Duchesne, Kaiwen Zhang, Ulrich Aivodji