AgentODE uses LLMs to discover ODE structures and infer parameter distributions from aggregate data, recovering consistent structures on benchmarks and RDEB clinical data with 231 observations from 46 patients.
hub
Advances in neural information processing systems , volume=
28 Pith papers cite this work. Polarity classification is still indexing.
28
Pith papers citing it
hub tools
citation-role summary
background 2
citation-polarity summary
years
2026 28roles
background 2representative citing papers
A GRPO-trained rubric generator that injects step-level criteria into ReAct agents improves deep-research benchmark scores over several test-time compute baselines.
AgentLens reveals 10.7% of passing SWE-agent trajectories exhibit Lucky Pass behaviors and introduces a process-level evaluation framework with a new annotated dataset of 1,815 trajectories.
Tool cloning is pervasive in agentic AI ecosystems, with 60% of high-Jaccard and 85% of high-ssdeep MCP repository pairs manually verified as true clones.
TacoMAS performs test-time co-evolution of agent capabilities and communication topology in LLM multi-agent systems via fast capability updates and slow meta-LLM topology edits, delivering 13.3% average gains over strong baselines on four benchmarks.
LLM agents encode tool necessity in pre-generation hidden states with high linear decodability (AUROC 0.89-0.96); Probe&Prefill uses this to reduce tool calls 48% with 1.7% accuracy loss.
MemCoE learns memory organization guidelines via contrastive feedback and then trains a guideline-aligned RL policy for memory updates, yielding consistent gains on personalization benchmarks.
Structural dependency graphs and staged pre-execution verification raise LLM-based EDA code pass rates to 82.5% (single-step) and 70-84% (multi-step) while halving tool calls by catching dependency violations before runtime.
Introduces MPT benchmark and PRefine method that models user preferences as evolving hypotheses to improve personalized tool calling accuracy with 1.24% of full-history token cost.
DeInfer reduces parallel inference communication cost for decomposed LLMs by up to 78% by moving collective operations into the low-rank latent space and redesigning KV-cache reconstruction for static graph compatibility.
Compiling repeated SOP nodes into environment-grounded, versioned tools cuts production p50 latency by 42% and end-to-end error rate by up to 53% in a 44-node fulfillment-center alarm-triage agent.
Introduces Active Task Driving Memory (ATMem) and STR-GRPO to move GUI agents from passive record storage to actively maintained task states, tested on a new mobile benchmark with progress and scope-aware metrics.
ISM maintains a self-refined bank of verified strategy schemas to enable continual mathematical reasoning in frozen LLMs, outperforming baselines on MATH-Hard and OlympiadBench while using 64-86% fewer schemas.
MemAudit combines counterfactual causal influence scores with memory consistency graphs to identify poisoned records in LLM agent memory, reducing MINJA attack success from 70% to 0% in QA and 83.3% to 0% in reasoning tasks.
PACE coordinates low-risk prompt evolution with validated higher-risk control-logic updates to improve frozen SLM agents on benchmarks without model retraining.
Partial harnesses for LLM agents, specifying only initial execution steps, achieve higher pass rates than fully decomposed workflows, as analyzed through trajectory alignment and validated in synthetic and terminal benchmarks.
ICRL uses joint RL training of solver and critic with distribution-calibration re-weighting and role-wise advantage estimation to internalize critique into unassisted LLM performance, yielding 6.4-point gains on agentic tasks and 7.0 on math reasoning with Qwen3 models.
Cochise is a 597-line reference harness with released replay/analysis tools and GOAD trajectory logs, intended to make LLM penetration-testing experiments reproducible and comparable.
ReFlect is a harness that wraps LLMs to detect and recover from reasoning errors, achieving 7-29 pp gains over direct CoT on long-horizon tasks and improving code patch quality to 82-87%.
Language models engage in counterexample-repair loops for conceptual definitions but produce increasingly verbose outputs without accuracy gains and hit diminishing returns quickly.
Janus is a method-agnostic plug-in that uses a Memory Momentum Trigger and compact hybrid evaluation to selectively accept LLM memory updates, yielding +2.7 to +4.6 accuracy gains over base updaters on six datasets.
Parallel WebBench reveals GRPO training raises web agent completion to 96% but leaves a large correctness gap from context-bound loops, premature termination, and synthesis collapse.
LiSA improves AI guardrails lifelong by inducing conservative policies from sparse noisy failure reports via structured memory, conflict-aware rules, and posterior lower-bound gating.
Skill1 trains a single RL policy to co-evolve skill selection, utilization, and distillation in language model agents from one task-outcome reward, using low-frequency trends to credit selection and high-frequency variation to credit distillation, outperforming baselines on ALFWorld and WebShop.
citing papers explorer
-
LLM-Guided ODE Discovery and Parameter Inference from Small-Cohort Aggregate Data
AgentODE uses LLMs to discover ODE structures and infer parameter distributions from aggregate data, recovering consistent structures on benchmarks and RDEB clinical data with 231 observations from 46 patients.
-
Co-ReAct: Rubrics as Step-Level Collaborators for ReAct Agents
A GRPO-trained rubric generator that injects step-level criteria into ReAct agents improves deep-research benchmark scores over several test-time compute baselines.
-
AgentLens: Revealing The Lucky Pass Problem in SWE-Agent Evaluation
AgentLens reveals 10.7% of passing SWE-agent trajectories exhibit Lucky Pass behaviors and introduces a process-level evaluation framework with a new annotated dataset of 1,815 trajectories.
-
Evaluating Tool Cloning in Agentic-AI Ecosystems
Tool cloning is pervasive in agentic AI ecosystems, with 60% of high-Jaccard and 85% of high-ssdeep MCP repository pairs manually verified as true clones.
-
TacoMAS: Test-Time Co-Evolution of Topology and Capability in LLM-based Multi-Agent Systems
TacoMAS performs test-time co-evolution of agent capabilities and communication topology in LLM multi-agent systems via fast capability updates and slow meta-LLM topology edits, delivering 13.3% average gains over strong baselines on four benchmarks.
-
LLM Agents Already Know When to Call Tools -- Even Without Reasoning
LLM agents encode tool necessity in pre-generation hidden states with high linear decodability (AUROC 0.89-0.96); Probe&Prefill uses this to reduce tool calls 48% with 1.7% accuracy loss.
-
Learning How and What to Memorize: Cognition-Inspired Two-Stage Optimization for Evolving Memory
MemCoE learns memory organization guidelines via contrastive feedback and then trains a guideline-aligned RL policy for memory updates, yielding consistent gains on personalization benchmarks.
-
Structural Verification for Reliable EDA Code Generation without Tool-in-the-Loop Debugging
Structural dependency graphs and staged pre-execution verification raise LLM-based EDA code pass rates to 82.5% (single-step) and 70-84% (multi-step) while halving tool calls by catching dependency violations before runtime.
-
Latent Preference Modeling for Cross-Session Personalized Tool Calling
Introduces MPT benchmark and PRefine method that models user preferences as evolving hypotheses to improve personalized tool calling accuracy with 1.24% of full-history token cost.
-
Co-evolving Agent Architectures and Interpretable Reasoning for Automated Optimization
DeInfer reduces parallel inference communication cost for decomposed LLMs by up to 78% by moving collective operations into the low-rank latent space and redesigning KV-cache reconstruction for static graph compatibility.
-
Tool-Making and Self-Evolving LLM Agents in Low-Latency Systems
Compiling repeated SOP nodes into environment-grounded, versioned tools cuts production p50 latency by 42% and end-to-end error rate by up to 53% in a 44-node fulfillment-center alarm-triage agent.
-
What Memory Do GUI Agents Really Need? From Passive Records to Active Task-Driving States
Introduces Active Task Driving Memory (ATMem) and STR-GRPO to move GUI agents from passive record storage to actively maintained task states, tested on a new mobile benchmark with progress and scope-aware metrics.
-
ISM:Self-Improving Strategy Memory for Continual Mathematical Reasoning
ISM maintains a self-refined bank of verified strategy schemas to enable continual mathematical reasoning in frozen LLMs, outperforming baselines on MATH-Hard and OlympiadBench while using 64-86% fewer schemas.
-
MemAudit: Post-hoc Auditing of Poisoned Agent Memory via Causal Attribution and Structural Anomaly Detection
MemAudit combines counterfactual causal influence scores with memory consistency graphs to identify poisoned records in LLM agent memory, reducing MINJA attack success from 70% to 0% in QA and 83.3% to 0% in reasoning tasks.
-
PACE: Two-Timescale Self-Evolution for Small Language Model Agents
PACE coordinates low-risk prompt evolution with validated higher-risk control-logic updates to improve frozen SLM agents on benchmarks without model retraining.
-
Harnesses for Inference-Time Alignment over Execution Trajectories
Partial harnesses for LLM agents, specifying only initial execution steps, achieve higher pass rates than fully decomposed workflows, as analyzed through trajectory alignment and validated in synthetic and terminal benchmarks.
-
ICRL: Learning to Internalize Self-Critique with Reinforcement Learning
ICRL uses joint RL training of solver and critic with distribution-calibration re-weighting and role-wise advantage estimation to internalize critique into unassisted LLM performance, yielding 6.4-point gains on agentic tasks and 7.0 on math reasoning with Qwen3 models.
-
Cochise: A Reference Harness for Autonomous Penetration Testing
Cochise is a 597-line reference harness with released replay/analysis tools and GOAD trajectory logs, intended to make LLM penetration-testing experiments reproducible and comparable.
-
ReFlect: An Effective Harness System for Complex Long-Horizon LLM Reasoning
ReFlect is a harness that wraps LLMs to detect and recover from reasoning errors, achieving 7-29 pp gains over direct CoT on long-horizon tasks and improving code patch quality to 82-87%.
-
The Counterexample Game: Iterated Conceptual Analysis and Repair in Language Models
Language models engage in counterexample-repair loops for conceptual definitions but produce increasingly verbose outputs without accuracy gains and hit diminishing returns quickly.
-
The Past Is Prologue: A Plug-in Controller for Selective Updates in Sequentially Evolving LLM Memory
Janus is a method-agnostic plug-in that uses a Memory Momentum Trigger and compact hybrid evaluation to selectively accept LLM memory updates, yielding +2.7 to +4.6 accuracy gains over base updaters on six datasets.
-
When Web Agents Finish but Still Fail: Reproducible Triggers and Trace Diagnostics for Parallel Web Exploration
Parallel WebBench reveals GRPO training raises web agent completion to 96% but leaves a large correctness gap from context-bound loops, premature termination, and synthesis collapse.
-
LiSA: Lifelong Safety Adaptation via Conservative Policy Induction
LiSA improves AI guardrails lifelong by inducing conservative policies from sparse noisy failure reports via structured memory, conflict-aware rules, and posterior lower-bound gating.
-
Skill1: Unified Evolution of Skill-Augmented Agents via Reinforcement Learning
Skill1 trains a single RL policy to co-evolve skill selection, utilization, and distillation in language model agents from one task-outcome reward, using low-frequency trends to credit selection and high-frequency variation to credit distillation, outperforming baselines on ALFWorld and WebShop.
-
BaRA: Budget-constrained and Reliable Web Data Collection Agent
BaRA improves valid link discovery and multimodal artifact extraction in budget-constrained web data collection via BFS liveness checks, rule-based validation, and self-reflection.
-
AEM: Adaptive Entropy Modulation for Multi-Turn Agentic Reinforcement Learning
AEM adaptively modulates response-level entropy in agentic RL to improve credit assignment and exploration-exploitation balance, yielding gains on ALFWorld, WebShop, and SWE-bench.
- EgoMemReason: A Memory-Driven Reasoning Benchmark for Long-Horizon Egocentric Video Understanding
- FlexSQL: Flexible Exploration and Execution Make Better Text-to-SQL Agents