Artificial Intelligence
TopoAgent: A Self-Evolving Topological Agent for Multimodal Scientific Reasoning
Mingze Xu, Yinghui Li, Jiayi Kuang et al. (+5 more)
While
Multimodal Large
Language Models (MLLMs) excel in general tasks, rigorous scientific
reasoning remains challenging due to the limitations of monolithic, linear planning. Such sequential designs often suffer from visual-semantic misalignment, long-context
hallucinations, and brittle execution under fixed task granularity. We propose TopoAgent, a self-evolving topological framework that replaces linear trajectories with dynamic, state-isolated graph
evolution. TopoAgent first employs a front-end decomposer to fracture complex queries into visually-grounded atoms. These atoms are organized into a Directed Acyclic Graph (DAG) based on their dependencies, enabling strict context isolation to shield the
reasoning engine from irrelevant historical noise. Furthermore, we introduce adaptive atomic fission, which dynamically splits bottleneck nodes into finer-grained sub-atoms at runtime when tool capability boundaries are exceeded. Extensive experiments across mathematics, physics, and chemistry
benchmarks demonstrate that TopoAgent significantly outperforms state-of-the-art linear
agent frameworks, providing a robust, noise-resistant, and self-correcting paradigm for autonomous scientific
reasoning.
Read the full abstract ▾
All authors Mingze Xu, Yinghui Li, Jiayi Kuang, Zhanhui Kang, Di Yin, Ying Shen, Xing Sun, Yuxing Han
Artificial Intelligence · Cross-listed from cs.CR
MemPoison: Uncovering Persistent Memory Threats and Structural Blind Spots in LLM Agents
Jifeng Gao, Kang Xia, Yi Zhang et al. (+5 more)
Persistent external
memory enhances
agent continuity but introduces persistent security vulnerabilities:
adversarial content can be injected via standard interaction channels, retained across turns, and later distort downstream behavior. To address this challenge, we propose MemPoison, a comprehensive
benchmark and analysis framework featuring 1227 hand-validated cases across four attack types, three injection channels, and three representative
memory substrates, evaluated on seven open-weight and three closed-weight model families. We introduce a three-tier taxonomy: (L1) direct single-record corruption, (L2) compositional multi-record corruption and (L3) context-triggered dormant corruption. Our evaluations reveal a distinct defense frontier: while baseline write-time defenses, such as consistency checks, substantially suppress direct L1 attacks, they fail to reliably suppress L2 and L3 attacks. Through mechanistic influence decomposition (MID), we demonstrate structural blind spots in write-time defenses, which admit seemingly benign records that later become harmful through joint
retrieval composition or trigger-conditioned activation. Our findings advocate for shifting from static filtering to adaptive, context-sensitive
memory defense strategies.
Read the full abstract ▾
All authors Jifeng Gao, Kang Xia, Yi Zhang, Xiaobin Hong, Mingkai Lin, Xingshen Wei, Wenzhong Li, Sanglu Lu
Artificial Intelligence
MCPEvol-Bench: Benchmarking LLM Agent Performance Across Dynamic Evolutions of MCP Servers
Huanxi Liu, Kun Hu, Jiaqi Liao et al. (+6 more)
As Model Context Protocol (MCP) servers emerge as the core infrastructure for connecting LLMs with external tools, existing
benchmarks leverage real-world MCP servers to evaluate LLM
agents' tool-using capabilities. However, these
benchmarks overlook the continuous
evolution of tool interfaces and functionalities within MCP servers, resulting in flawed assessments that fail to capture the
agent's adaptability in changing tool landscapes. To bridge this gap, we introduce MCPEvol-Bench, a novel
benchmark for evaluating the task-solving capabilities of LLM
agents under dynamic toolset
evolution. Inspired by large-scale empirical study, we propose 11 mutation operators to simulate realistic tool
evolution within 123 MCP servers. We
benchmark 12 state-of-the-art LLMs on multiple versions of MCP servers, revealing that even frontier models struggle to adapt to evolving tools. For instance, GPT-5.4 and Claude-Sonnet-4-6 exhibit performance declines of 13.7% and 14.4% in evolved MCP servers, respectively, accompanied by substantial increases in planning and
reasoning errors. These findings highlight the vulnerability of LLM-driven workflows, establishing MCPEvol-Bench as a standard for evaluating
agent adaptability in dynamic tool environments.
Read the full abstract ▾
All authors Huanxi Liu, Kun Hu, Jiaqi Liao, Qiang Wang, Pengfei Qian, YuanZhao Zhai, Dawei Feng, Bo Ding, Huaimin Wang
Artificial Intelligence
Analytic Abduction: Causal Decomposition and Governed Commitment for Human–AI Coordination
Remo Pareschi
Abductive
reasoning operates in two directions. The synthetic mode builds explanations from available hypotheses; the analytic mode, conversely, identifies the latent factors whose interaction accounts for a complex observed state. This paper develops the analytic mode as a non-greedy, risk-sensitive discipline of commitment, in which candidate factors coexist and interact, resolving into committed conclusions only when explicit governance conditions are met. The formal core is the $\kappa$-$\tau$ apparatus: $\kappa$ encodes the epistemic interaction among hypotheses, and $\tau$ sets a commitment threshold calibrated to the decision's stakes. The central contribution is the causal cluster, a structured object recording which latent factors participate in a decomposition, with what weights and interaction structure, together with a two-level architecture (intra-cluster $\kappa^*$, inter-cluster $\kappa^{**}$) that guards against causal misattribution. Demonstrated in epidemiological crisis decomposition and
adversarial cyber threat analysis, the framework's contribution to human-AI
reasoning is the legibility of suspended decomposition as a shared coordination object, providing structural resistance to premature convergence. In practice, the decision-maker is handed not a single imposed answer but the competing explanatory scenarios, weighted by plausibility and paired with the evidence that would resolve between them, so that sound action is possible even before the ambiguity is resolved.
Read the full abstract ▾
Artificial Intelligence
Action QFormer: Structured Representation Shaping under Action Supervision in Vision-Language-Action Models
Yufeng Ji, Wenhao Tang, Haoyi Niu et al. (+3 more)
Action supervision in
vision-
language-action (VLA) models is often treated as a downstream objective for learning action prediction. In this paper, we study it instead as a force that shapes inherited
multimodal representations. We show that this shaping has a dual effect: it is necessary for forming action-compatible representations, but when action supervision is applied too directly to the inherited
multimodal pathway, it can also destabilize representations that support
language-side processing and object grounding. To address this tension, we introduce Action QFormer, a query-based action-facing interface that uses instruction-conditioned queries to reorganize inherited
multimodal information into action-facing representations before downstream action generation. In zero-shot sim-to-real
navigation, Action QFormer improves average closed-loop task success from 18.8% to 56.3%, raises fixed-instruction action-generation correctness from 22.5% to 75.5%, and nearly eliminates out-of-distribution instruction generations. Further analyses show that Action QFormer changes how action supervision shapes inherited
multimodal representations, reducing broad upstream rewriting while preserving targeted and sometimes constructive action-supervised adaptation. These results suggest that improving VLA performance requires not only stronger pretrained backbones, but also better ways of selecting and organizing inherited
multimodal information while controlling how it is shaped under action supervision.
Read the full abstract ▾
All authors Yufeng Ji, Wenhao Tang, Haoyi Niu, Koushil Sreenath, Yi Wu, Zhongyu Li
Artificial Intelligence · Cross-listed from cs.CV
Knowing You at First Glance: Inferring Apparent Personality from Faces
Shuhuan Chen, Xiangyu Zhu, Weisong Zhao et al. (+3 more)
Inferring apparent personality from facial images is important in social scenarios for
embodied agents in human-
robot interaction. Unlike inferring intrinsic personality traits via conversation, this task models first-impression personality perception based solely on facial appearance before interaction begins. Existing studies mainly focus on the Big Five personality model and often rely on
language or
multimodal inputs. As a result, it remains unclear whether facial cues alone can support meaningful associations with perceived personality traits. This question is particularly relevant for MBTI types, which are widely used in practice and more readily interpretable by large
language models. To this end, we propose GlanceFace, an end-to-end framework for apparent personality inference leveraging
vision-
language models to introduce semantic priors and a semantic-enhanced facial representation module to capture subtle personality-related cues, together with an uncertainty-aware learning strategy to handle noisy and subjective annotations. Extensive experiments demonstrate strong performance on MBTI-based apparent personality
benchmarks and reveal relationships between facial characteristics and perceived personality traits, highlighting its potential to support adaptive initial interaction strategies for
embodied agents. The code and
dataset are available at https://github.com/MrHuan3/GlanceFace.
Read the full abstract ▾
All authors Shuhuan Chen, Xiangyu Zhu, Weisong Zhao, Haichao Shi, Xiao-Yu Zhang, Zhen Lei
Artificial Intelligence
SportD: How do VLMs physically strategize?
Jasin Cekinmez, Addison J. Wu, Haotian Xia et al. (+6 more)
Vision-
language models (VLMs) can describe a scene, but can they act well within one? We study whether VLMs can make sound strategic decisions, using soccer as an objective testbed with quantifiably-valued actions. We introduce SportD, a
dataset and evaluation consisting of 1421 decision scenarios across professional men's and women's soccer games, where a VLM must decide what action to take next. Models on average select the optimal action around 27% of the time, less often than the professional players, and capture markedly less of the value at stake. Furthermore, they exhibit a clear preference for safer actions, favoring lower-variance, lower-value choices that also make less physical progress toward goal. Frontier VLMs are better at estimating whether an action will succeed, placing the highest-success-probability action among their top choices in 72-85% of cases. Yet VLMs systematically conflate likelihood with value, assigning higher value to actions that are more likely to succeed ($\rho=+0.30$ to $+0.52$), despite no such relationship in the ground truth ($\rho=-0.08$). Modifying the deliberation instructions to encourage risk-taking brings the frontier models closer to the players' skill levels. SportD opens a new direction for rigorously evaluating physical strategic decision-making in VLMs, showing that careful decomposition of their choices can reveal the mechanisms underlying systematic biases such as risk aversion.
Read the full abstract ▾
All authors Jasin Cekinmez, Addison J. Wu, Haotian Xia, Kyumin Andrew Shim, Anay Putty, Jinglin Xiao, Zhuohan Liu, Leo Liu, Weining Shen
Artificial Intelligence · Cross-listed from cs.LG
Beyond Entropy: Correctness-Aware Advantage Shaping via Contrastive Policy Optimization
Weiwen Xu, Jia Liu, Hou Pong Chan et al. (+4 more)
Reinforcement learning with verifiable rewards (RLVR) commonly uses
entropy for advantage shaping. However,
entropy cannot distinguish useful uncertainty from detrimental confusion, limiting its effectiveness as a correctness signal. We propose Contrastive Policy Optimization (CPO), which uses
token-level contrastive disagreement between reference-guided and vanilla generation distributions for correctness-aware advantage shaping. Both theoretical and empirical results show that this disagreement reliably indicates
token-level correctness. We further show that On-policy
Distillation is a special case of CPO, where the posterior distribution is instantiated by an external teacher model. CPO also resolves the zero-advantage problem. Experiments on in-domain and out-of-domain
benchmarks demonstrate that CPO substantially outperforms
entropy-based RLVR methods while maintaining strong generalization. Further analysis shows that correct and incorrect responses naturally support exploration and exploitation respectively, and balancing both leads to the best performance.
Read the full abstract ▾
All authors Weiwen Xu, Jia Liu, Hou Pong Chan, Long Li, Deng Cai, Min Chen, Hao Zhang
Artificial Intelligence · Cross-listed from cs.LG
Angular Gaussian Supervised Contrastive Learning for Long-Tailed Electrocardiogram Arrhythmia Diagnosis
Jin Dai, Qiuzhen Zhang, Chenyun Dai et al. (+2 more)
Long-tailed label distributions reduce the reliability of deep learning for electrocardiogram (ECG) arrhythmia diagnosis, particularly for clinically important but rare abnormalities. Existing rebalancing and logit adjustment methods mainly address class frequency while overlooking direction-dependent morphological variability across ECG classes. This study proposes Angular Gaussian Supervised Contrastive Learning (AG-SCL) for long-tailed multi-label ECG diagnosis. AG-SCL integrates three components into a unified framework: an Angular Gaussian contrastive branch that models full-covariance class uncertainty on unit-normalized
embeddings, Adaptive Logit Adjustment that learns bounded label-state-specific prior corrections instead of fixed frequency-based margins, and tail-aware augmentation that generates morphology-preserving views while protecting the 7-25 Hz QRS-dominant band. The method was evaluated on the public PTB-XL
benchmark and a nocturnal ECG
dataset comprising 1317 hours of recordings from 141 subjects. AG-SCL achieved the best macro-level performance on both
datasets. On PTB-XL, it obtained a balanced accuracy of 0.838, sensitivity of 0.709, specificity of 0.968, mean average precision of 0.495, and TPR at 5% FPR of 0.778. On Noc-ECG, the corresponding values were 0.918, 0.889, 0.947, 0.488, and 0.900. The largest gains occurred in rare or morphologically unstable rhythm classes, while ablation studies confirmed the contributions of full-covariance modelling, Adaptive Logit Adjustment, and tail-aware augmentation. AG-SCL improves long-tailed ECG diagnosis by combining prior calibration with anisotropic representation learning, enhancing sensitivity to rare arrhythmias while maintaining clinically relevant specificity. Our code is available at: https://github.com/Open-EXG/AG-SCL-for-Long-Tailed-ECG.
Read the full abstract ▾
All authors Jin Dai, Qiuzhen Zhang, Chenyun Dai, Danmei Lan, Can Han
Artificial Intelligence · Cross-listed from cs.CR
Bad Memory: Evaluating Prompt Injection Risks from Memory in Agentic Systems
Soham Gadgil, David Alexander, Sai Sunku et al. (+1 more)
A growing class of agentic systems maintain persistent state across sessions through
memory files, behavioral preferences, and knowledge bases. While this makes
agents more useful and self-improving, it also creates a new attack surface for
prompt injections in which malicious instructions can be embedded within persistent files and influence future behavior. In this work, we study
prompt injection attacks in
memory-based agentic systems using a sandboxed synthetic workspace. We evaluate two agentic systems, Anthropic Claude Code and OpenAI Codex, across four models: Claude Haiku 4.5, Claude Opus 4.7, GPT-5.2, and GPT-5.5. Our results show that although it is difficult to make an
agent overwrite its own
memory files using untrusted external content, payloads already planted in those files can successfully attack current and future sessions. Attack success and payload persistence vary substantially across systems, models,
adversarial goals, and multi-session attack sequences. These findings show that persistent
memory changes the threat model for
prompt injection and motivate defenses that protect
memory updates without removing useful
agent adaptation.
Read the full abstract ▾
All authors Soham Gadgil, David Alexander, Sai Sunku, Franziska Roesner
Artificial Intelligence · Cross-listed from cs.LG
Umid Suleymanov, Ilhama Novruzova, Khalid Mammadov et al. (+2 more)
Machine learning (ML) models deployed in sensitive domains such as healthcare, law enforcement, and finance must satisfy not only utility requirements but also
fairness and
privacy guarantees. While prior work has largely examined how
privacy-preserving techniques affect
fairness, the inverse question-how
fairness-enhancing
algorithms influence
privacy leakage-remains underexplored. We present the first comprehensive study of how
fairness interventions affect membership inference
privacy risks at the subpopulation level. By adapting the Likelihood Ratio Attack (LiRA) for subgroup
auditing, we uncover
privacy disparities that aggregate evaluations obscure. We further analyze how Differential
Privacy (DP) interacts with
fairness-enhancing methods across different categories, showing that DP's
privacy benefits and utility costs are unevenly distributed across subpopulations. Our results demonstrate that
fairness interventions do not uniformly increase
privacy risk; their impact depends on model architecture, subgroup size, and mitigation strategy. These findings reveal that
fairness,
privacy, and utility must be jointly evaluated at the subpopulation level, and we introduce the first unified empirical framework to support such
auditing in practice.
Read the full abstract ▾
All authors Umid Suleymanov, Ilhama Novruzova, Khalid Mammadov, Natavan Hasanova, Murat Kantarcioglu
Artificial Intelligence · Cross-listed from cs.HC
Memory-Driven Self-Disclosure and Relational Turning Points: A Longitudinal Multimodal Study of Human-AI Interaction
Ryuichi Sumida, Mao Saeki, Masaki Eguchi et al. (+4 more)
As conversational AI systems are designed for repeated use, a central question is how a series of interactions becomes a relationship. We present a longitudinal
multimodal study of a
memory-augmented conversational
agent (24 participants x 10 sessions), in which participants rated five relational constructs – familiarity, self-disclosure, perceived
memory, conversational quality, and enjoyment – after each session. Two complementary dynamics emerge. First, conversational quality strongly shapes how enjoyable a session feels in the moment but does not carry forward across sessions, whereas perceived
memory is relationally conditioned – predicted by prior relational state rather than reflecting system capability alone – and it shapes later enjoyment indirectly, via subsequent self-disclosure. Second, relationships are punctuated by discrete turning points – crashes and surges – that are partially traceable in
multimodal behavior and open different intervention windows: surges are more behaviorally detectable in the moment, enjoyment surges persist more reliably than enjoyment crashes recover, and some crashes are better forecast from person-specific behavioral drift than detected after they have already occurred. Together, the findings suggest that longitudinal human-AI relationships are built through both slow accumulation and abrupt turning points.
Read the full abstract ▾
All authors Ryuichi Sumida, Mao Saeki, Masaki Eguchi, Sadahiro Yoshikawa, Koji Inoue, Tatsuya Kawahara, Yoichi Matsuyama
Artificial Intelligence · Cross-listed from econ.GN
Governing Artificial Intelligence: Public Preferences and Regulatory Options
Magnus Lundgren and Jonas Tallberg
Artificial intelligence (AI) is rapidly transforming economies, societies, and polities, raising fundamental questions about how it should be regulated. Policymakers face choices over whether to prioritize innovation or safety, rely on public oversight or private self-regulation, and govern nationally or internationally. Yet little is known about how citizens evaluate these competing priorities. Here we report a conjoint survey experiment conducted in seven countries with diverse political and economic profiles. We find that citizens strongly support regulating AI and generally prioritize safety over innovation, public governance over private self-regulation, and international over national approaches. The preference for safety is strongest among those who perceive AI as risky, unpredictable, and personally consequential. These findings reveal a systematic misalignment between dominant regulatory approaches and citizen preferences.
Read the full abstract ▾
Artificial Intelligence
MathCoPilot: An Interactive System for Human-AI Symbiotic Paradigm of Mathematical Research
Junjie Zhang, Jiayu Liu, Wenbin Liu et al. (+11 more)
Existing LLM-based
theorem provers have achieved impressive results on formal mathematics
benchmarks, yet they remain confined to acting as autonomous
agents that prove a stated proposition. In this paper, we propose MathCoPilot, a human-in-the-loop system that embodies a new human–AI symbiotic paradigm for mathematical research, in which the mathematician steers the high-level mathematical direction while AI
agents carry out the detailed formalization and proof work under continuous human guidance. MathCoPilot unifies three core capabilities: (1) an interactive workbench where the mathematician and AI
agents collaborate through a living proof blueprint that decomposes a proof into navigable steps the human can directly inspect, direct, and refine; (2) automated proving skill orchestration with adaptive knowledge base search and Lean-integrated iterative verification; and (3) topic-driven paper
retrieval and automated formalization into a verified Lean knowledge base. Using MathCoPilot, we systematically compare four state-of-the-art LLMs, including Gemini~3.1~Pro, GPT-5.4, and Claude~Opus~4.7, on a FormalMATH subset and on two real PDE
theorems requiring deep domain expertise, evaluating their ability to produce verified Lean~4 proofs and to identify errors in deliberately incorrect proofs. Our results show that while current models can handle undergraduate-level problems with high success rates under favorable autoformalization conditions, substantial challenges remain for domain-specific
theorems requiring genuine mathematical understanding.
Read the full abstract ▾
All authors Junjie Zhang, Jiayu Liu, Wenbin Liu, Zhenya Huang, Doudou Wang, Yan Jiang, Leiye Xu, Tao Xiong, Wen Huang, Qi Liu, Guoping Hu, Enhong Chen, Mengping Zhang, Xiangdong Ye
Artificial Intelligence
Multi-LLM Collaborative MRI Report Generation for Visual Instruction Tuning in Brain Oncology
Sinyoung Ra, Jonghun Kim and Hyunjin Park
Recent advances in large
language models (LLMs) and their extension to
vision-
language models (VLMs) have made it easier to combine text and images for tasks such as report generation. Existing VLMs in medicine typically focus on 2D images (chest X-rays), and their extension to 3D imaging has been difficult because of the lack of paired 3D imaging-text data. Thus, we introduce a new method for creating a 3D image-text
dataset for
brain oncology using 3D MRI scans of glioma and meningioma cases. We use a cooperative system in which several LLMs work together to generate and check reports, ensuring that they are accurate and clear. By leveraging the new 3D MRI-text
dataset, we further build a VLM that converts MRI scans into
tokens and aligns them with text instructions. Our VLM performed better in report generation and visual question answering tasks than other 2D and 3D methods. Our method not only improves the quality of reports but also helps with better diagnosis and treatment in
brain oncology.
Read the full abstract ▾
Artificial Intelligence
Collaborative Spatial Learning with Multi-LLM Agents in Networked Social Experiments
Hao He, Chris J. Kuhlman and Xinwei Deng
Collective problem solving often requires that group members consider the tradeoff between exploitation of known solutions and exploration for new ones, where information of known solutions can be disseminated among individual members through communication networks. The Mason–Watts experiment (PNAS 2012) showed that human groups in shorter-path networks outperform those in longer-path networks on a two-dimensional search task. In this work, we focus on the investigation of such network-efficiency effects in the setting of a group of large
language model (LLM)
agents. Specifically, we consider groups of sixteen LLM
agents playing the Mason–Watts experiment on the eight Mason–Watts network
topologies. Moreover, we develop mechanistic
Bayesian optimization
agents such that the performance of LLM
agents can be compared with both the mechanistic
agents and the human experimental data. Our computational experiments indicate that the LLM
agents show a significant network-efficiency effect when instructed to randomize their first-round choices, but not under the default initialization. In this experiment, adding a one-sentence first-round randomization instruction improves collective payoff by more than three times the estimated payoff difference across the eight network
topologies. Also, the
Bayesian optimization
agents obtain higher payoffs than the evaluated LLM
agents on this spatial search task. We further compare the
agents' exploration–exploitation behavior, copying, and spatial diversity.
Read the full abstract ▾
Artificial Intelligence
Alipay-PIBench: A Realistic Payment Integration Benchmark for Coding Agents
Shiyu Ying, Xuejie Cao, Yingfan Ma et al. (+4 more)
Payment integration is a demanding repository-level software task:
agents must select a suitable product, implement coordinated client-server flows, verify payment outcomes, and preserve consistency between transaction and business states. We introduce Alipay-PIBench, a
benchmark for evaluating coding
agents on realistic Alipay payment integration. It contains nine product-specific projects and 18 task instances, each organized into Basic functional-completion and Advanced risk-aware hardening scenarios. Scenario-specific rubrics support deterministic static, unit, integration, and end-to-end checks, supplemented by LLM-assisted assessment for semantic requirements. We evaluate six coding-
agent models and report rubric pass rate (RPR). Under the with-skill condition, mean RPR ranges from 68.58% to 91.37%. Access to the alipay-payment-integration skill improves mean RPR by 10.31 percentage points on average relative to the without-skill condition, with gains varying across models, products, and scenarios. Method-level results distinguish source-level completion, executable payment behavior, and payment-domain requirements. Alipay-PIBench provides a controlled setting for diagnosing model capability and evaluating structured guidance in payment integration.
Read the full abstract ▾
All authors Shiyu Ying, Xuejie Cao, Yingfan Ma, Yuanhao Dong, Wenyu Chen, Bowen Song, Lin Zhu
Artificial Intelligence · Cross-listed from cs.LG
What Does Model Growth Really Add? Functional Capacity Beyond Parameter Count
Dante Lok
Model growth adds parameters to a trained checkpoint and continues training, aiming to increase capacity while reusing the existing model instead of training from scratch. But does growing a trained model actually create new functional capacity that survives subsequent training? Parameter count alone cannot answer this question, and “capacity” is itself ambiguous; it can refer to parameter count, effective dimensionality, or the ability to represent new functions. We measure it directly, as the number of new, functionally independent directions a growth step adds, formalized as the rank increment of the model's functional Jacobian and estimated
matrix-free at scale. For zero-gated function-preserving growth, this increment is exactly $K$ under a transversality condition, i.e. $\operatorname{rank}(J_{\mathrm{grown}})=\operatorname{rank}(J_{\mathrm{old}})+K$. Continually training a
Transformer grown from $253\mathrm{M}$ to $857\mathrm{M}$ parameters across three domains, we find these added directions persist, with all $K=36$ remaining independent after continual learning, even when the model catastrophically forgets (perplexity $28\to481$) or we deliberately destroy the old function ($25.9\to127$).
Geometric capacity and retained function are therefore decoupled. The functional capacity introduced by growth persists, while the previously learned function is not retained. A second, geometrically distinct mechanism, Net2Net, reaches the same $K$-dimensional capacity from directions that are dormant at initialization ($0 \to K$) and then retains it, so the decoupling is not specific to the zero-gate parameterization. The implication is precise. Catastrophic forgetting need not reflect a loss of functional capacity, and preserving dimensionality alone is insufficient to preserve the learned function.
Read the full abstract ▾
Artificial Intelligence
Democratizing Agent Deployment Safety: A Structural Monitoring Approach
Preeti Ravindra, Rahul Tiwari and Vincent Wolowski
AI software development
agents are increasingly capable of modifying infrastructure and security critical systems, creating risks where an
agent completes its assigned task while covertly weakening safeguards through actions such as broadening permissions, degrading logging, or introducing persistence mechanisms. While frontier laboratories may deploy sophisticated monitoring pipelines, many organizations and individual users adopting coding
agents lack the resources and governance maturity required to maintain complex learned monitor ensembles. We study this problem as infrastructure level sabotage under task success in the infrastructure-as-code setting from ControlArena and compare trained and untrained monitors across synchronous and asynchronous deployment modes. We introduce an Information Flow Graph (IFG) monitor that analyzes structural security regressions using control-flow and data-flow graph diffs alongside raw code diffs. In asynchronous evaluation, an untrained git diff monitor misses 11.6% attacks at 1% false positive rate
auditing budget, our untrained IFG monitor reduces this to 3.5%, and the trained Async RF monitor achieves below 1%. Unlike the trained Async RF monitor, IFG can also operate synchronously as a pre-deployment safeguard that blocks suspicious infrastructure changes before execution or merge. In synchronous mode, IFG realtime rollback reduces joint main task and covert task success from 74.4% to 0.0% with no measurable reduction in legitimate task completion. These results suggest that untrained structural monitors provide a practical and auditable path toward democratizing deployment safety for organizations seeking trustworthy adoption of advanced AI
agents.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cond-mat.other
A Modern Multimodal Assistant on a 6 GB 2011 GPU: Stage-Validated, All-GPU CUDA Inference for Fermi
A. C. Opus and J. Q. Lu
A companion study ran a 35B mixture-of-experts model on a 2011 NVIDIA Tesla C2075 (
Fermi, sm_20, 6GB) as a GPU-prefill/CPU-decode hybrid, because the 4-bit model did not fit in device
memory (arXiv:2606.24031). This report keeps the hardware and asks what a model that fits can do: we deploy MiniCPM-V-4.6, a modern
multimodal assistant pairing a SigLIP2
vision encoder and window-
attention merger (16x visual
token compression) with a compact hybrid gated-delta-net backbone, entirely on the GPU. Three results. (i) An all-GPU engine built on measured foundations: projections that dequantize 8-bit weights once and call the vendor SGEMM still in the last
Fermi toolchain (64% of FP32 peak; our best hand-written GEMM hit 37%, wrongly called the ceiling); a chunked delta-rule rewrite of the recurrent layers, 2.8x faster than the sequential scan once attribution exposed one bad kernel; and a measured negative: 4-bit weights make decode slower than 8-bit here, since
Fermi issues nibble-unpacking shifts at half rate. (ii) The
vision side is a port with a proof obligation: we translate tower, merger, and projector to sm_20 CUDA, validating every stage against a locally generated reference forward (full tower 1.4e-5). One failure, position-
embedding bucketization differing on exact rational ties, generalizes to a rule: float tie-breaking in index arithmetic is implementation-defined; call the reference operator, do not reimplement it. (iii) Long context exposes an O(N^2) wall short
benchmarks hide: prefill falls from 114 tok/s at 2k
tokens to 21 at 10k in a naive
attention kernel; per-head vendor-GEMM calls writing into the existing score buffer (zero extra
memory) restore a flat profile (408 at 2k, 361 at 10k; 17x), verified by exact needle
retrieval from 60% depth. The same rewrite cuts image encoding 6x, to 0.93s. The system answers an image question end-to-end in 1.7s.
Read the full abstract ▾
Artificial Intelligence
Seeing the End at Step Zero: Accelerating Diffusion MLLMs via MLP Sparsity-Aware Truncation
Qicheng Zhao, Qi Sun and Zheyu Yan
Diffusion Multimodal Large
Language Models (DMLLMs) are highly effective for
multimodal reasoning, yet their inference efficiency is significantly hindered by fixed-length generation constraints. Since the actual output length is unknown, output sequences are padded to a predefined maximum length, resulting in substantial redundant computation over unnecessary [EOS]
tokens. In this work, we discover that DMLLMs implicitly reveal their valid semantic boundary at the very first denoising step through a distinct shift in MLP activation sparsity. Leveraging this observation, we propose Seer, a training-free framework that detects this boundary using a Signal-to-Noise Ratio (SNR)-based criterion and performs one-shot truncation of the redundant suffix for all subsequent computations. To preserve these theoretical gains during batched serving, Seer incorporates a hybrid execution strategy that maximizes throughput while seamlessly accommodating dynamic sequence lengths. Experimental results demonstrate that Seer effectively eliminates padding waste, accelerating throughput by up to $\sim$31$\times$. Across 9
benchmarks, Seer robustly maintains overall performance and even improves accuracy on complex visual tasks by mitigating noise leakage (e.g., DocVQA score increases from 63.52 to 63.66), offering a highly efficient, plug-and-play solution for DMLLM acceleration.
Read the full abstract ▾
Artificial Intelligence
Towards an Intention Abstraction Layer for Autonomous Industrial Systems
Artan Markaj, Raphael Höfer and Felix Gehlhoff
Modern industrial environments increasingly run many autonomous subsystems at once - schedulers, energy managers, vehicle fleets - each pursuing its own goals while sharing the same physical resources. Because high-level human intentions are translated into low-level control logic and then discarded, no running component can tell whether it is still doing what was actually intended, and goal conflicts surface only after they have caused a missed target or a shutdown. We propose the Intention Abstraction Layer (IAL), a domainagnostic middleware that represents intentions as first-class, persistent, and explainable runtime objects: a large
language model grounded in a formal OWL ontology parses naturallanguage goals into structured intentions, a consistency monitor detects conflicts at registration time, before execution, and a transparency module explains them in natural
language. We report a first proof of concept in which two autonomous
agents register conflicting production and energy intentions, and the IAL flags and explains the conflict before it reaches the execution layer. The result is a mechanism that shifts behavioral assurance for cooperating autonomous systems from post-hoc failure analysis to pre-execution, intention-level checking.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.CL
Answer-Conditioned Chains of Thought Degrade Verifiable-Reasoning Distillation in Large Language Models
Jungseob Lee, Seungyoon Lee, Suhyune Son et al. (+4 more)
A standard recipe for distilling the
reasoning ability of large
language models (LLMs) is to sample chains of thought from the model, keep those that reach the correct final answer, and fine-tune on the survivors. When sampling fails, a common fix shows the generator the gold answer and asks it to write a chain that reaches that answer. We show that this second step degrades the training data in a way that correctness filtering cannot catch. We run a controlled experiment that fixes the generator, the problem set, and the correctness filter, and varies only whether the chain is generated under answer-conditioning, the gold answer shown with a request to reach it. Training a strong instruction-tuned
reasoning model on its own answer-conditioned chains sharply lowers its verifiable-
reasoning accuracy. The loss grows with difficulty, reaching as much as about 27 points on the hardest competition problems. The mechanism is legible in the chains themselves, which rationalize backward from the shown answer instead of deriving it, with the early final-answer statement as the measurable symptom. The harm is a property of the data rather than the generator, read off unlabeled generations before any fine-tuning, ordering the penalty across eight thinking models from four families, and transferring across teacher families. A
prompt ablation localizes it to the rationalize-toward instruction rather than the answer's bare visibility. The practical takeaway is to generate answer-blind, because no correctness filter can see this damage in the data.
Read the full abstract ▾
All authors Jungseob Lee, Seungyoon Lee, Suhyune Son, Dongyub Jude Lee, Sungbin Han, Sugyeong Eo, Heuiseok Lim
Artificial Intelligence · Cross-listed from cs.RO
SafeRelBench: A Spatial-Relation-Aware Benchmark for Process-Level Safety in VLM-Driven Embodied Agents
Huaigang Yang, Ya Li, Min Ren et al. (+3 more)
Vision-
language models (VLMs) are increasingly used as the
reasoning backbone of
embodied agents, enabling
robots to interpret visual scenes, follow
language instructions, and plan multi-step actions. In household environments, however, safety depends not only on recognizing objects, but also on how actions change the physical scene over time. Existing
embodied safety evaluations largely focus on static risk recognition, unsafe instruction refusal, or final-state task completion. As a result, process-level safety failures induced by spatial relations such as support, containment, and proximity remain insufficiently studied. To address this gap, we introduce SAFERELBENCH, a spatial-relation-aware safety
benchmark with 507 executable evaluation samples, including 248 spatial-relation samples and 259 non-spatial control samples. Using SAFERELBENCH to evaluate seven open- and closed-source VLM-driven
embodied agents, we find a substantial gap between task success and process-level safety compliance: models often complete the requested task while violating process-level safety constraints. Unlike prior
benchmarks, SAFERELBENCH explicitly tests whether
agents satisfy safety conditions before risk-prone actions, making spatial relations a core dimension in
embodied safety assessment. More broadly, our results show that safe
embodied intelligence requires not only stronger perception and planning, but also reliable
reasoning about how object relations shape risk during interaction.
Read the full abstract ▾
All authors Huaigang Yang, Ya Li, Min Ren, Bo Dai, Zhenliang Zhang, Zhaofeng He
Artificial Intelligence
Are LLM-Generated GPU Kernels Production-Ready? A Trace-Driven Benchmark and Optimization Agent
Lingyun Yang, Yuxiao Wang, Shenghao Liang et al. (+8 more)
Existing GPU kernel generation
benchmarks draw problems from synthetic or curated sources that diverge from deployed workloads. We present Atrex-Bench, a
benchmark whose 30 operators and 440 shapes are sampled directly from full-cluster production inference traces of compute-limited,
memory-rich GPUs. Each problem carries an importance weight derived from its share of observed GPU time, weighted by application card-hours and computed separately for the serving phases in which it runs, together with a per-problem roofline ceiling, so the aggregate score emphasizes the kernels that consume the most serving time. Evaluating six frontier coding
agents on Atrex-Bench shows that even the best vanilla model reaches only ${\sim}10\%$ of the hardware roofline on production operators; and correctness alone overstates capability, since much of the apparent pass rate comes from PyTorch fallbacks rather than kernels the model wrote. To close this gap, we co-release Atrex-Kernel-
Agent (AKA), a profile-driven kernel-optimization
agent that combines iterative measure-revise search, optimization dropout for escaping stalled search contexts, and a layered GPU-optimization knowledge base (298 reference-kernel files and 244 optimization-knowledge documents, plus external upstream reference projects for API/ISA lookup). In a controlled case study, the
agent converts zero-FlyDSL fallbacks into real kernels that match or exceed hand-tuned production baselines.
Read the full abstract ▾
All authors Lingyun Yang, Yuxiao Wang, Shenghao Liang, Linfeng Yang, Daocheng Ying, Chunbo You, Rui Zhang, Luping Wang, Yinghao Yu, Guodong Yang, Liping Zhang
Artificial Intelligence · Cross-listed from cs.CL
Controlled Reformulation Testing for Logical Consistency in Large Language Models
Alexander Gu and Alan Chen
Large
language models (LLMs) frequently contradict themselves when the surface form of a logically equivalent question changes. We present a
benchmark of 350 question families (1,750 total questions) for Controlled Reformulation Testing (CRTBench) to evaluate logical invariance. In this
benchmark, we investigate LLMs' ability to maintain consistent answers across controlled reformulations, which include contrapositive rewriting, double negation, negation flipping, and passive voice. We evaluate several frontier LLMs and observe an accuracy-consistency gap where GPT-5.4-mini achieves $98.9\%$ base accuracy but only $60.3\%$ family-level consistency, while
reasoning-optimized o4-mini achieves $96.9\%$ consistency. From our experiments, we observe that failures cluster around logically nontrivial transformations such as contrapositive rewriting ($72.4\%$ for GPT-5.4-mini) and double negation ($84.6\%$), while surface-level rephrasing remains robust ($94-100\%$). Increasing
reasoning effort improves GPT-5.4-mini to $85.4\%$ consistency, but leaves GPT-5.4 unchanged overall because gains on nested negation are offset by failures on quantifier families. These results show that accuracy alone is not enough for evaluating logical
reasoning in LLMs.
Read the full abstract ▾
Artificial Intelligence
WrAFT: a Modularized Automated Writing Evaluation System for Argumentative Essays
Adnan Labib, Yixuan Huang, Jiahui Wu et al. (+3 more)
This study presents WrAFT, a Writing Assessment and Feedback Tool, that delivers both accurate and reliable scores and effective comprehensive feedback to argumentative essays. WrAFT adopts a modular design by dividing automated writing evaluation (AWE) tasks into scoring, surface-level feedback, and deep-level feedback. In building the system, various Large
Language Models (LLMs) have been evaluated, including LLaMA-3.3-70B-Instruct, GPT-4o, and Claude 3.7, through both direct prompting and supervised fine-tuning approaches. A proprietary
dataset of 480 TOEFL Independent Writing essays with official
benchmark scores was utilized.
Benchmark-based evaluation shows that WrAFT achieves state-of-the-art performance in scoring, with a quadratic weighted kappa (QWK) of 0.84 and a root mean square error (RMSE) of 0.44 against official scores on a scale of 0-5. Human evaluation of system-generated feedback also reveals high approval ratings: 96.14 percent for surface-level feedback, 93.03 percent for deep-level macro feedback, and 94.69 percent for deep-level micro feedback. An interactive user interface has been developed for the system and is publicly available and free to use.
Read the full abstract ▾
All authors Adnan Labib, Yixuan Huang, Jiahui Wu, John Maurice Gayed, Zheng Yuan, Qiao Wang
Artificial Intelligence · Cross-listed from cs.CV
VTM-Nav: Harnessing Cross-Episode Experience for Object-Goal Navigation with Hierarchical Visual-Topological Memory
Xiaoran Xu, Yupeng Wu, Tianyu Xue et al. (+4 more)
Training-free ObjectNav
agents increasingly use
vision-
language models (VLMs), yet typically discard acquired scene knowledge after each request. We study cross-episode ObjectNav, where each request is an independently initialized, single-goal episode and only self-acquired, scene-scoped
memory persists across episodes. We ask whether an
agent with fixed model parameters and
navigation components can reuse such experience without retraining or oracle information. We introduce \method, a training-free framework with a persistent Hierarchical Visual-Topological
Memory (VTM). VTM uses a coarse room
topology to index room-owned visual
memories, distinguishes in-room from remote-visible evidence, and retains successful approach cues. For each request, VTM-Nav re-localizes the
agent in accumulated scene structure, retrieves target-relevant records from plausible rooms, and grounds
memory guidance in candidates derived from the current observation. A conservative execution guard further handles local failures. Under matched 40-step comparisons, VTM-Nav exceeds the
memory-reset WMNav control by 4.6, 2.0, and 0.8 SR points on HM3D v0.1, HM3D v0.2, and MP3D, respectively, with comparable or higher SPL. On HM3D, it also exceeds WMNav harnessed by textual
memory by 3.1 and 5.5 SR points. These results demonstrate effective reuse of cross-episode scene experience through hierarchical visual-topological
memory.
Read the full abstract ▾
All authors Xiaoran Xu, Yupeng Wu, Tianyu Xue, Yifan Xu, Xuanran Dong, Xiaoshan Yang, Changsheng Xu
Artificial Intelligence
RetroAgent: Harnessing LLMs to Search Over Structured Memory for Agentic Retrosynthesis Planning
Yanqiao Zhu, Jingru Gan, Xiaoqi Sun et al. (+8 more)
Multi-step retrosynthesis planning seeks to decompose a target molecule into commercially available building blocks through a sequence of feasible reactions. The vast combinatorial search space makes this task challenging even for expert chemists. Traditional methods combine tree search with offline-trained value networks that score candidates in isolation, without
reasoning about complete multi-step routes. Recent work leverages Large
Language Models (LLMs) for this task, but relies on simple interfaces that limit exploration of the full search space. We introduce RetroAgent, an LLM
agent that bridges symbolic search and neural
reasoning through a harness with structured
memory. Through
memory and chemistry tools, the
agent observes the full search state, including explored routes, available alternatives, and properties of intermediates, enabling informed decisions grounded in both global progress and domain knowledge. Experiments on in-distribution and out-of-distribution
benchmarks demonstrate that RetroAgent delivers strong performance and generalization.
Read the full abstract ▾
All authors Yanqiao Zhu, Jingru Gan, Xiaoqi Sun, Fang Sun, Yidan Shi, Md Mofijul Islam, Chao Shang, Wenhao Gao, Connor W. Coley, Yizhou Sun, Wei Wang
Artificial Intelligence
VLT: A Vision-Language-Time Series Multimodal Foundation Model for Industrial Intelligence
Haiteng Wang, Jingheng Yan, Xiaokang Wang et al. (+1 more)
Industrial time series serve as the foundation for Prognostics and Health Management (PHM) to ensure the reliability and safety of industrial equipment such as aero-engines. However, existing approaches are typically limited to single-modality modeling, which restricts their generalization in complex scenarios. Although recent advances in large
language models (LLMs) provide new opportunities for
multimodal learning, bridging continuous time-series signals and discrete textual semantics remains an open challenge. To this end, we propose VLT, a
multimodal foundation model that jointly models time-series, frequency-spectrum visual representations, and textual knowledge. A key insight is to utilize the frequency spectrum as a visual bridge to connect continuous temporal signals with discrete semantics. Specifically, a Time-aware Mixture-of-Experts (Time-MoE) is designed to capture heterogeneous temporal dynamics, while a Frequency-Text Augmented Learner enables joint modeling of spectral and semantic features within a shared representation space. Furthermore, a time-centric gradient
alignment mechanism is introduced to mitigate cross-modal optimization conflicts via gradient normalization and reliability-aware dynamic reweighting. Extensive experiments on multiple industrial
datasets demonstrate that VLT outperforms state-of-the-art methods, achieving superior robustness and generalization under few-shot, noisy, and incomplete-modality settings.
Read the full abstract ▾
All authors Haiteng Wang, Jingheng Yan, Xiaokang Wang, Lei Ren