EvoVuln evolves executable detection policies for five smart-contract vulnerability types using cold-start synthetic testing followed by few-shot refinement on five vulnerable and five safe contracts, reaching 71% macro F1 and enabling a small model to beat a large zero-shot model by 19 points at un
hub
Chain-of-thought prompting elicits reasoning in large language models
42 Pith papers cite this work. Polarity classification is still indexing.
42
Pith papers citing it
hub tools
citation-role summary
background 3
citation-polarity summary
roles
background 3polarities
background 3representative citing papers
CodeChat-Eval shows LLMs lose 19.2% to 69.2% functional correctness over multi-turn refinement dialogues, with largest drops on logic-level and additive changes.
LatentOmni proposes a latent-space cross-modal reasoning framework that uses feature-level supervision and Omni-Sync Position Embedding to align and synchronize audio-visual latents, supported by a new 35K interleaved reasoning dataset and showing gains over text CoT baselines.
CAN-QA creates 33,128 QA pairs from CAN traffic logs in 10 categories to test LLMs, which capture patterns but struggle with temporal reasoning and multi-condition inference.
WORC improves multi-agent LLM reasoning to 82.2% average accuracy by predicting and compensating for the weakest agent via targeted extra sampling rather than uniform reinforcement.
CoT prompting in LLM4Code shows mixed robustness that depends on model family, task structure, and perturbations destabilizing structural anchors, leading to trajectory deformations like lengthening, branching, and simplification.
D³ETOR combines debate-enhanced pseudo labeling from SAM with frequency-aware progressive debiasing in FADeNet to achieve state-of-the-art weakly-supervised camouflaged object detection using scribbles.
CARLA-GS is a modular pipeline that uses an LLM for semantic trajectory planning, CARLA for physics execution, and 3D Gaussian Splatting for photorealistic rendering to synthesize autonomous driving corner cases.
ConCise is a training-free protocol that compresses multi-step RAG context from O(N²) to O(N) tokens using conclusion chains and fused generation, achieving 64.63% average token savings.
ProCrit proposes a Proposal-Critic framework that synthesizes process-level annotations via agentic rollout and uses draft-critique-revise with mutual-refinement RL to improve multimodal sarcasm detection.
RCLAgent uses multi-agent recursion-of-thought with parallel reasoning on trace graphs to outperform prior methods in root cause localization accuracy and efficiency for microservice systems.
A bidirectional semantic complementary tool retrieval method using planning-based query enhancement and dynamic tool dependency graphs with neighborhood aggregation improves retrieval accuracy on remote sensing and general tool tasks.
EVOREC integrates locate-then-edit model editing with FA-constrained decoding to improve LLM-based service recommendation under evolution, reporting 25.9% average relative gain in Recall@5 over baselines and 22.3% over fine-tuning in dynamic scenarios.
Case study of CMBAgent on 18 astrophysical tasks finds strong performance on well-specified problems but frequent silent failures yielding physically inconsistent outputs.
Behavior Forest decouples multi-constraint travel planning into parallel behavior trees with LLM nodes and global coordination, yielding 6.67% and 11.82% gains over prior methods on two benchmarks.
CoDA aligns cross-domain latent reasoning representations in LLMs via CoT distillation and MMD to enable effective knowledge transfer without in-domain demonstrations.
Proactive multi-window state triggering plus Set-of-Mark alignment and multimodal LLM reasoning detects GUI defects in Android apps, reporting 184% more text truncation, 87.2% F1 on occlusion, and 40 defect-prone apps at 10% FPR.
PARM adapts reward models to multi-stage LLM pipelines via pipeline data and direct preference optimization, improving execution rate and solving accuracy on optimization benchmarks and showing transfer to GSM8K.
AQPIM performs in-memory product quantization of activations for LLMs on PIM hardware, reducing GPU-CPU communication by 90-98.5% and delivering 3.4x speedup over prior PIM methods.
LoReC enhances LLMs for graph tasks via attention redistribution, graph re-injection into FFN, and logit rectification, yielding improvements over GraphLLM and GNN baselines on diverse datasets.
Hallucinations in multimodal reasoners cluster at high-entropy pivot tokens due to intermediate-layer visual anchoring failure, and V-STAR mitigates this with attention rewards plus forced reflection.
A clinician-validated 46-item checklist and zero-shot LLM baselines audit discharge-summary completeness on 50 MIMIC-IV notes, with best models reaching only moderate agreement (κ≈0.5).
Chat-Scene++ improves 3D scene understanding in multimodal LLMs by representing scenes as context-rich object sequences with identifier tokens and grounded chain-of-thought reasoning, reaching state-of-the-art on five benchmarks using pre-trained encoders.
KG-Hopper uses RL to embed full multi-hop KG traversal and backtracking into a single LLM inference round, enabling a 7B model to outperform larger multi-step systems and compete with GPT-3.5/GPT-4o-mini on eight benchmarks.
citing papers explorer
-
Knowledge Over Parameters: Evolving Smart Contract Vulnerability Detection
EvoVuln evolves executable detection policies for five smart-contract vulnerability types using cold-start synthetic testing followed by few-shot refinement on five vulnerable and five safe contracts, reaching 71% macro F1 and enabling a small model to beat a large zero-shot model by 19 points at un
-
CodeChat-Eval: Evaluating Large Language Models in Multi-Turn Code Refinement Dialogues
CodeChat-Eval shows LLMs lose 19.2% to 69.2% functional correctness over multi-turn refinement dialogues, with largest drops on logic-level and additive changes.
-
LatentOmni: Rethinking Omni-Modal Understanding via Unified Audio-Visual Latent Reasoning
LatentOmni proposes a latent-space cross-modal reasoning framework that uses feature-level supervision and Omni-Sync Position Embedding to align and synchronize audio-visual latents, supported by a new 35K interleaved reasoning dataset and showing gains over text CoT baselines.
-
CAN-QA: A Question-Answering Benchmark for Reasoning over In-Vehicle CAN Traffic
CAN-QA creates 33,128 QA pairs from CAN traffic logs in 10 categories to test LLMs, which capture patterns but struggle with temporal reasoning and multi-condition inference.
-
Weak-Link Optimization for Multi-Agent Reasoning and Collaboration
WORC improves multi-agent LLM reasoning to 82.2% average accuracy by predicting and compensating for the weakest agent via targeted extra sampling rather than uniform reinforcement.
-
Structural Anchors and Reasoning Fragility:Understanding CoT Robustness in LLM4Code
CoT prompting in LLM4Code shows mixed robustness that depends on model family, task structure, and perturbations destabilizing structural anchors, leading to trajectory deformations like lengthening, branching, and simplification.
-
Debate-Enhanced Pseudo Labeling and Frequency-Aware Progressive Debiasing for Weakly-Supervised Camouflaged Object Detection with Scribble Annotations
D³ETOR combines debate-enhanced pseudo labeling from SAM with frequency-aware progressive debiasing in FADeNet to achieve state-of-the-art weakly-supervised camouflaged object detection using scribbles.
-
CARLA-GS: Decoupling Representation, Reasoning, and Physics Simulation for Autonomous Driving Corner-Case Synthesis
CARLA-GS is a modular pipeline that uses an LLM for semantic trajectory planning, CARLA for physics execution, and 3D Gaussian Splatting for photorealistic rendering to synthesize autonomous driving corner cases.
-
ConCise: Training-Free Conclusion-Chain State Compression for Cost-Efficient Multi-Step RAG Services
ConCise is a training-free protocol that compresses multi-step RAG context from O(N²) to O(N) tokens using conclusion chains and fused generation, achieving 64.63% average token savings.
-
ProCrit: Self-Elicited Multi-Perspective Reasoning with Critic-Guided Revision for Multimodal Sarcasm Detection
ProCrit proposes a Proposal-Critic framework that synthesizes process-level annotations via agentic rollout and uses draft-critique-revise with mutual-refinement RL to improve multimodal sarcasm detection.
-
Towards In-Depth Root Cause Localization for Microservices with Multi-Agent Recursion-of-Thought
RCLAgent uses multi-agent recursion-of-thought with parallel reasoning on trace graphs to outperform prior methods in root cause localization accuracy and efficiency for microservice systems.
-
Bidirectional Semantic Complementary Tool Retrieval for Remote Sensing Agents
A bidirectional semantic complementary tool retrieval method using planning-based query enhancement and dynamic tool dependency graphs with neighborhood aggregation improves retrieval accuracy on remote sensing and general tool tasks.
-
When Model Editing Meets Service Evolution: A Knowledge-Update Perspective for Service Recommendation
EVOREC integrates locate-then-edit model editing with FA-constrained decoding to improve LLM-based service recommendation under evolution, reporting 25.9% average relative gain in Recall@5 over baselines and 22.3% over fine-tuning in dynamic scenarios.
-
Plausible but Wrong: A case study on Agentic Failures in Astrophysical Workflows
Case study of CMBAgent on 18 astrophysical tasks finds strong performance on well-specified problems but frequent silent failures yielding physically inconsistent outputs.
-
Decoupled Travel Planning with Behavior Forest
Behavior Forest decouples multi-constraint travel planning into parallel behavior trees with LLM nodes and global coordination, yielding 6.67% and 11.82% gains over prior methods on two benchmarks.
-
CoDA: Towards Effective Cross-domain Knowledge Transfer via CoT-guided Domain Adaptation
CoDA aligns cross-domain latent reasoning representations in LLMs via CoT distillation and MMD to enable effective knowledge transfer without in-domain demonstrations.
-
Proactive Detection of GUI Defects in Multi-Window Scenarios via Multimodal Reasoning
Proactive multi-window state triggering plus Set-of-Mark alignment and multimodal LLM reasoning detects GUI defects in Android apps, reporting 184% more text truncation, 87.2% F1 on occlusion, and 40 defect-prone apps at 10% FPR.
-
PARM: Pipeline-Adapted Reward Model
PARM adapts reward models to multi-stage LLM pipelines via pipeline data and direct preference optimization, improving execution rate and solving accuracy on optimization benchmarks and showing transfer to GSM8K.
-
AQPIM: Breaking the PIM Capacity Wall for LLMs with In-Memory Activation Quantization
AQPIM performs in-memory product quantization of activations for LLMs on PIM hardware, reducing GPU-CPU communication by 90-98.5% and delivering 3.4x speedup over prior PIM methods.
-
LoReC: Rethinking Large Language Models for Graph Data Analysis
LoReC enhances LLMs for graph tasks via attention redistribution, graph re-injection into FFN, and logit rectification, yielding improvements over GraphLLM and GNN baselines on diverse datasets.
-
Cognitive Pivot Points and Visual Anchoring: Unveiling and Rectifying Hallucinations in Multimodal Reasoning Models
Hallucinations in multimodal reasoners cluster at high-entropy pivot tokens due to intermediate-layer visual anchoring failure, and V-STAR mitigates this with attention rewards plus forced reflection.
-
CareTransition-Audit: A Benchmark to Audit Discharge Summaries for Efficient Care Transitions
A clinician-validated 46-item checklist and zero-shot LLM baselines audit discharge-summary completeness on 50 MIMIC-IV notes, with best models reaching only moderate agreement (κ≈0.5).
-
Chat-Scene++: Exploiting Context-Rich Object Identification for 3D LLM
Chat-Scene++ improves 3D scene understanding in multimodal LLMs by representing scenes as context-rich object sequences with identifier tokens and grounded chain-of-thought reasoning, reaching state-of-the-art on five benchmarks using pre-trained encoders.
-
KG-Hopper: Empowering Compact Open LLMs with Knowledge Graph Reasoning via Reinforcement Learning
KG-Hopper uses RL to embed full multi-hop KG traversal and backtracking into a single LLM inference round, enabling a 7B model to outperform larger multi-step systems and compete with GPT-3.5/GPT-4o-mini on eight benchmarks.
-
TDA-RC: Task-Driven Alignment for Knowledge-Based Reasoning Chains in Large Language Models
TDA-RC embeds topological patterns from multi-round reasoning into CoT via persistent homology and a repair agent, yielding better accuracy-efficiency trade-offs than ToT or GoT on tested datasets.
-
C2F-Thinker: Coarse-to-Fine Reasoning with Hint-Guided Reinforcement Learning for Multimodal Sentiment Analysis
C2F-Thinker combines structured coarse-to-fine chain-of-thought reasoning with hint-guided GRPO reinforcement learning to achieve competitive fine-grained sentiment regression and superior cross-domain generalization in multimodal analysis.
-
VERDI: VLM-Embedded Reasoning for Autonomous Driving
VERDI aligns perception, prediction, and planning outputs of end-to-end AD models with VLM-generated text features at training time to embed structured reasoning, yielding up to 11% better l2 distance and 10% higher non-collision rate in closed-loop tests.
-
General Hazard Detection
Introduces CompliVision dataset and active learning framework for rule-based hazard compliance assessment using vision-language models grounded in safety standards.
-
Agentic AIs Are the Missing Paradigm for Out-of-Distribution Generalization in Foundation Models
Agentic AI systems are required to overcome the parameter coverage ceiling that prevents foundation models from handling certain out-of-distribution cases.
-
AutoVQA-G: Self-Improving Agentic Framework for Automated Visual Question Answering and Grounding Annotation
AutoVQA-G is a self-improving framework that generates VQA-G datasets with higher visual grounding accuracy than leading multimodal LLMs via iterative CoT verification and prompt refinement.
-
Rethinking the Necessity of Adaptive Retrieval-Augmented Generation through the Lens of Adaptive Listwise Ranking
AdaRankLLM shows adaptive listwise reranking outperforms fixed-depth retrieval for most LLMs by acting as a noise filter for weak models and an efficiency optimizer for strong ones, with lower context use.
-
Cross-Lingual Attention Distillation with Personality-Informed Generative Augmentation for Multilingual Personality Recognition
ADAM uses personality-guided LLM augmentation and cross-lingual attention distillation to raise balanced accuracy on multilingual personality recognition to 0.6332 on Essays and 0.7448 on Kaggle, outperforming standard BCE loss.
-
Can LLMs Make (Personalized) Access Control Decisions?
LLMs reflect users' privacy preferences in access control decisions with up to 86% agreement and can promote safer behavior, but personalization trades off higher individual match for potentially less secure results when users over-permission.
-
OpsAgent: An Evolving Multi-agent System for Incident Management in Microservices
OpsAgent presents a training-free multi-agent framework with dual self-evolution for automated incident management in microservices, claiming SOTA results on OPENRCA benchmark and successful production deployment at Lenovo.
-
RF Instrument Agent (RFIA): Empowering RF Instruments with Natural Language Understanding, Scheduling and Execution of Complex Tasks
RFIA presents a decoupled LLM-agent architecture for natural-language RF instrument control, with a structured knowledge base and hybrid execution graphs, evaluated on a 16-task VNA benchmark.
-
DeliCIR: Memory-Guided Test-Time Deliberation via Multi-Agent Collaboration for Composed Image Retrieval
A hierarchical multi-agent framework that fuses three CLIP-based retrieval views and then applies tournament-style test-time reasoning reports the best published scores on CIRR, CIRCO, and FashionIQ.
-
Hierarchical Prompting with Dual LLM Modules for Robotic Task and Motion Planning
A dual-LLM hierarchical framework for robotic task and motion planning, integrating object detection, achieves 86% success across 24 test scenarios ranging from simple spatial commands to infeasible requests.
-
CyberAId: AI-Driven Cybersecurity for Financial Service Providers
CyberAId is a proposed on-premise multi-agent system that coordinates LLM subagents with classical security tools to improve threat response and regulatory alignment in financial services.
-
Rethinking Wireless Communications through Formal Mathematical AI Reasoning
Proposes a three-layer framework using formal AI reasoning for verification, derivation, and discovery in wireless communications theory.
-
AICCE: AI Driven Compliance Checker Engine
AICCE combines RAG-based retrieval of protocol specs with dual LLM pipelines for debate-driven explanations or fast script execution, reporting up to 99% accuracy on IPv6 samples.
-
From Prompts to Pavement: LMMs-based Agentic Behavior-Tree Generation Framework for Autonomous Vehicles
An agentic LLM/LVM framework generates adaptive behavior trees on-the-fly for AV navigation in CARLA+Nav2 simulation, succeeding in obstacle avoidance where static BTs fail.
-
LEO-RobotAgent: A General-purpose Robotic Agent for Language-driven Embodied Operator
LEO-RobotAgent is a general-purpose framework that enables LLMs to independently plan, use tools, and collaborate with humans while operating multiple robot types for unpredictable tasks.