MemGym unifies agent gyms into a memory benchmark with isolated scoring across tool-use, research, coding, and computer-use regimes plus a lightweight reward model for tractable coding evaluation.
hub Canonical reference
Reflexion: Language agents with verbal reinforcement learning.Advances in neural information processing systems, 36:8634–8652
Canonical reference. 75% of citing Pith papers cite this work as background.
44
Pith papers citing it
Background
75% of classified citations
hub tools
citation-role summary
background 15
baseline 5
citation-polarity summary
years
2026 44representative citing papers
SMMBench is a benchmark evaluating multimodal agents on cross-source reasoning, conflict resolution, preference reasoning, and action prediction, showing current systems struggle with evidence distributed across heterogeneous sources.
MemDocAgent generates consistent hierarchical repository-level code documentation by combining dependency-aware traversal with memory-guided agent interactions that accumulate work traces.
QueST adapts LLMs at test time by generating query-specific problem-solution pairs for self-supervised fine-tuning, improving reasoning performance without external data.
SDP constructs a task-induced state space from raw text by having agents commit to and certify natural-language predicates as states, enabling structured planning and analysis in unstructured language environments.
SkillSmith is a boundary-first compiler-runtime system that turns skill packages into minimal executable interfaces, cutting token usage 57%, thinking iterations 43%, and solve time 51% versus raw skill injection on SkillsBench.
ReCrit frames critic interaction as a correctness-transition problem and uses quadrant-based RL rewards to improve LLM performance on scientific reasoning benchmarks by rewarding corrections and robustness while penalizing sycophancy.
MAGE uses a four-subgraph co-evolutionary knowledge graph plus dual bandits to externalize and retrieve experience for stable self-evolution of frozen language-model agents, showing gains on nine diverse benchmarks.
EquiMem calibrates shared memory in multi-agent debate by computing a game-theoretic equilibrium from agent queries and paths, outperforming heuristics and LLM validators across benchmarks while remaining robust to adversarial agents.
SkillGuard extracts executable environment contracts from LLM skill documents to detect only relevant drifts, reporting zero false positives on 599 cases, 100% precision in known-drift tests, and raising one-round repair success from 10% to 78%.
AgentForesight introduces an online auditor model that predicts decisive errors in multi-agent trajectories at the earliest step using a coarse-to-fine reinforcement learning recipe on a new curated dataset AFTraj-2K.
RewardHarness self-evolves a tool-and-skill library from 100 preference examples to reach 47.4% accuracy on image-edit evaluation, beating GPT-5, and yields stronger RL-tuned models.
MemCompiler reframes memory use as state-conditioned compilation, delivering relevant guidance via text and latent channels to improve embodied agent performance up to 129% and cut latency 60% versus static injection.
Think-with-Rubrics has LLMs generate rubrics internally before responding, outperforming external rubric-as-reward baselines by 3.87 points on average across benchmarks.
A VOI-based controller for dual inference budgets improves multi-hop QA performance by prioritizing search actions and selectively finalizing answers.
Priority ranking offers a low-cost direct evaluation for harness optimizers that correlates with their real multi-step optimization performance, supported by the Shor dataset of 182 scenarios.
Maestro uses outcome-based RL to train a lightweight policy that orchestrates ensembles of frozen expert models and skills, reporting 70.1% average accuracy across ten multimodal benchmarks and outperforming GPT-5 and Gemini-2.5-Pro while generalizing to unseen components.
Auto-Dreamer trains an offline memory consolidator via GRPO on agent performance to abstract cross-session patterns, outperforming baselines by 7 points on ScienceWorld with 12x smaller memory and generalizing to ALFWorld and WebArena.
MementoGUI introduces a modular memory-control framework with working and episodic memory operators that improves long-horizon GUI agent performance over history-replay and text-only baselines.
A hybrid LLM-RL red teaming framework generates adaptive attack campaigns in simulated enterprise networks to evaluate the robustness of AI-enabled SOAR systems.
DrugSAGE accumulates cross-task memory of skills, statistical evidence, and recurring errors to let LLM agents achieve top-ranked performance on molecular property prediction tasks with reduced or zero test-time search.
MAP improves LLM agent reasoning by constructing a structured cognitive map of the environment before task execution, yielding performance gains on benchmarks like ARC-AGI-3 and superior training data via the new MAP-2K dataset.
RESD turns failure trajectories into token-level supervision via retrospective reflections and a persistent global playbook, enabling faster improvement than standard self-distillation or GRPO with only one rollout per prompt.
FATE lets LLM agents self-evolve safer behaviors by generating and filtering repairs from their own failure trajectories using verifiers and Pareto optimization.
citing papers explorer
-
MemGym: a Long-Horizon Memory Environment for LLM Agents
MemGym unifies agent gyms into a memory benchmark with isolated scoring across tool-use, research, coding, and computer-use regimes plus a lightweight reward model for tractable coding evaluation.
-
SMMBench: A Benchmark for Source-Distributed Multimodal Agent Memory
SMMBench is a benchmark evaluating multimodal agents on cross-source reasoning, conflict resolution, preference reasoning, and action prediction, showing current systems struggle with evidence distributed across heterogeneous sources.
-
Remember Your Trace: Memory-Guided Long-Horizon Agentic Framework for Consistent and Hierarchical Repository-Level Code Documentation
MemDocAgent generates consistent hierarchical repository-level code documentation by combining dependency-aware traversal with memory-guided agent interactions that accumulate work traces.
-
Query-Conditioned Test-Time Self-Training for Large Language Models
QueST adapts LLMs at test time by generating query-specific problem-solution pairs for self-supervised fine-tuning, improving reasoning performance without external data.
-
State-Centric Decision Process
SDP constructs a task-induced state space from raw text by having agents commit to and certify natural-language predicates as states, enabling structured planning and analysis in unstructured language environments.
-
SkillSmith: Compiling Agent Skills into Boundary-Guided Runtime Interfaces
SkillSmith is a boundary-first compiler-runtime system that turns skill packages into minimal executable interfaces, cutting token usage 57%, thinking iterations 43%, and solve time 51% versus raw skill injection on SkillsBench.
-
ReCrit: Transition-Aware Reinforcement Learning for Scientific Critic Reasoning
ReCrit frames critic interaction as a correctness-transition problem and uses quadrant-based RL rewards to improve LLM performance on scientific reasoning benchmarks by rewarding corrections and robustness while penalizing sycophancy.
-
MAGE: Multi-Agent Self-Evolution with Co-Evolutionary Knowledge Graphs
MAGE uses a four-subgraph co-evolutionary knowledge graph plus dual bandits to externalize and retrieve experience for stable self-evolution of frozen language-model agents, showing gains on nine diverse benchmarks.
-
EquiMem: Calibrating Shared Memory in Multi-Agent Debate via Game-Theoretic Equilibrium
EquiMem calibrates shared memory in multi-agent debate by computing a game-theoretic equilibrium from agent queries and paths, outperforming heuristics and LLM validators across benchmarks while remaining robust to adversarial agents.
-
Skill Drift Is Contract Violation: Proactive Maintenance for LLM Agent Skill Libraries
SkillGuard extracts executable environment contracts from LLM skill documents to detect only relevant drifts, reporting zero false positives on 599 cases, 100% precision in known-drift tests, and raising one-round repair success from 10% to 78%.
-
AgentForesight: Online Auditing for Early Failure Prediction in Multi-Agent Systems
AgentForesight introduces an online auditor model that predicts decisive errors in multi-agent trajectories at the earliest step using a coarse-to-fine reinforcement learning recipe on a new curated dataset AFTraj-2K.
-
RewardHarness: Self-Evolving Agentic Post-Training
RewardHarness self-evolves a tool-and-skill library from 100 preference examples to reach 47.4% accuracy on image-edit evaluation, beating GPT-5, and yields stronger RL-tuned models.
-
MemCompiler: Compile, Don't Inject -- State-Conditioned Memory for Embodied Agents
MemCompiler reframes memory use as state-conditioned compilation, delivering relevant guidance via text and latent channels to improve embodied agent performance up to 129% and cut latency 60% versus static injection.
-
Think-with-Rubrics: From External Evaluator to Internal Reasoning Guidance
Think-with-Rubrics has LLMs generate rubrics internally before responding, outperforming external rubric-as-reward baselines by 3.87 points on average across benchmarks.
-
Inference-Time Budget Control for LLM Search Agents
A VOI-based controller for dual inference budgets improves multi-hop QA performance by prioritizing search actions and selectively finalizing answers.
-
Towards Direct Evaluation of Harness Optimizers via Priority Ranking
Priority ranking offers a low-cost direct evaluation for harness optimizers that correlates with their real multi-step optimization performance, supported by the Shor dataset of 182 scenarios.
-
Maestro: Reinforcement Learning to Orchestrate Hierarchical Model-Skill Ensembles
Maestro uses outcome-based RL to train a lightweight policy that orchestrates ensembles of frozen expert models and skills, reporting 70.1% average accuracy across ten multimodal benchmarks and outperforming GPT-5 and Gemini-2.5-Pro while generalizing to unseen components.
-
Auto-Dreamer: Learning Offline Memory Consolidation for Language Agents
Auto-Dreamer trains an offline memory consolidator via GRPO on agent performance to abstract cross-session patterns, outperforming baselines by 7 points on ScienceWorld with 12x smaller memory and generalizing to ALFWorld and WebArena.
-
MementoGUI: Learning Agentic Multimodal Memory Control for Long-Horizon GUI Agents
MementoGUI introduces a modular memory-control framework with working and episodic memory operators that improves long-horizon GUI agent performance over history-replay and text-only baselines.
-
A Red Teaming Framework for Evaluating Robustness of AI-enabled Security Orchestration, Automation, and Response Systems
A hybrid LLM-RL red teaming framework generates adaptive attack campaigns in simulated enterprise networks to evaluate the robustness of AI-enabled SOAR systems.
-
DrugSAGE:Self-evolving Agent Experience for Efficient State-of-the-Art Drug Discovery
DrugSAGE accumulates cross-task memory of skills, statistical evidence, and recurring errors to let LLM agents achieve top-ranked performance on molecular property prediction tasks with reduced or zero test-time search.
-
MAP: A Map-then-Act Paradigm for Long-Horizon Interactive Agent Reasoning
MAP improves LLM agent reasoning by constructing a structured cognitive map of the environment before task execution, yielding performance gains on benchmarks like ARC-AGI-3 and superior training data via the new MAP-2K dataset.
-
Learning with Rare Success but Rich Feedback via Reflection-Enhanced Self-Distillation
RESD turns failure trajectories into token-level supervision via retrospective reflections and a persistent global playbook, enabling faster improvement than standard self-distillation or GRPO with only one rollout per prompt.
-
On-Policy Self-Evolution via Failure Trajectories for Agentic Safety Alignment
FATE lets LLM agents self-evolve safer behaviors by generating and filtering repairs from their own failure trajectories using verifiers and Pareto optimization.
-
STAR: Failure-Aware Markovian Routing for Multi-Agent Spatiotemporal Reasoning
STAR presents a failure-aware routing framework using a state-conditioned transition policy and an agent routing matrix combining expert routes with learned recoveries from execution traces to improve multi-agent spatiotemporal reasoning.
-
The Trap of Trajectory: Towards Understanding and Mitigating Spurious Correlations in Agentic Memory
Agentic memory improves clean reasoning but worsens performance when spurious patterns are present in stored trajectories; CAMEL calibration reduces this reliance while preserving clean performance.
-
From History to State: Constant-Context Skill Learning for LLM Agents
Constant-context skill learning trains reusable task-family modules for LLM agents using a deterministic state block for progress tracking and subgoal rewards, achieving 89.6% unseen success on ALFWorld, 76.8% on WebShop, and 66.4% on SciWorld with Qwen3-8B while reducing prompt tokens 2-7x.
-
Agent-World: Scaling Real-World Environment Synthesis for Evolving General Agent Intelligence
Agent-World autonomously synthesizes verifiable real-world tasks and uses continuous self-evolution to train 8B and 14B agents that outperform proprietary models on 23 benchmarks.
-
Verify Before You Fix: Agentic Execution Grounding for Trustworthy Cross-Language Code Analysis
A framework combining universal AST normalization, hybrid graph-LLM embeddings, and strict execution-grounded validation achieves 89-92% intra-language accuracy and 74-80% cross-language F1 while resolving 70% of vulnerabilities at 12% failure rate.
-
Do Agent Societies Develop Intellectual Elites? The Hidden Power Laws of Collective Cognition in LLM Multi-Agent Systems
LLM agent societies develop power-law coordination cascades and intellectual elites through an integration bottleneck that grows with system size.
-
What Do Agents Communicate? Characterizing Information Exchange in Multi-Agent Systems
Systematic study of inter-agent communication in LLM multi-agent systems shows reasoning and verification are critical for performance, with a new augmentation technique recovering 86.2% of failures.
-
Code as Agent Harness
A survey that organizes existing work on LLM-based agents around code as the central harness, structured in three layers of interfaces, mechanisms, and multi-agent scaling, with applications across domains and listed open challenges.
-
Robo-Cortex: A Self-Evolving Embodied Agent via Dual-Grain Cognitive Memory and Autonomous Knowledge Induction
Robo-Cortex proposes a self-evolving embodied navigation agent using dual-grain cognitive memory and autonomous knowledge induction from trajectories, reporting SPL gains on IGNav, AR, AEQA and preliminary real-robot tests.
-
Beyond Scaling: Agents Are Heading to the Edge
Personal agents require edge deployment to preserve high-fidelity local context and zero-latency loops, as claimed through three structural shifts away from cloud-centric designs.
-
SkillsVote: Lifecycle Governance of Agent Skills from Collection, Recommendation to Evolution
SkillsVote is a governance system for agent skills that profiles corpora, recommends via search, and gates updates on successful reusable outcomes, yielding benchmark gains without model changes.
-
EGL-SCA: Structural Credit Assignment for Co-Evolving Instructions and Tools in Graph Reasoning Agents
EGL-SCA co-evolves instructions and tools via structural credit assignment in graph reasoning agents and reports 92% average success on four benchmarks.
-
M2A: Synergizing Mathematical and Agentic Reasoning in Large Language Models
M2A uses null-space model merging to combine mathematical and agentic reasoning in LLMs, raising SWE-Bench Verified performance from 44.0% to 51.2% on Qwen3-8B without retraining.
-
Bian Que: An Agentic Framework with Flexible Skill Arrangement for Online System Operations
Bian Que is an agentic framework using a unified operational paradigm, flexible Skill Arrangement, and self-evolving mechanism to automate O&M tasks, achieving 75% alert reduction and over 50% MTTR cut in production deployment.
-
Evo-MedAgent: Beyond One-Shot Diagnosis with Agents That Remember, Reflect, and Improve
Evo-MedAgent adds three evolving memory stores to LLM agents for chest X-ray diagnosis, raising MCQ accuracy from 0.68 to 0.79 on GPT-5-mini and 0.76 to 0.87 on Gemini-3 Flash without any training.
-
SpaceMind: A Modular and Self-Evolving Embodied Vision-Language Agent Framework for Autonomous On-orbit Servicing
SpaceMind is a self-evolving modular VLM agent framework that achieves 90-100% navigation success in nominal conditions and recovers from failures via experience distillation, with zero-code transfer to physical robots for on-orbit tasks.
-
"Theater of Mind" for LLMs: A Cognitive Architecture Based on Global Workspace Theory
Global Workspace Agents (GWA) is proposed as an active, event-driven cognitive architecture for LLMs featuring an entropy-based intrinsic drive and dual-layer memory to enable sustained self-directed agency.
-
Skills-Coach: A Self-Evolving Skill Optimizer via Training-Free GRPO
Skills-Coach optimizes LLM agent skills via task generation, prompt/code tuning, comparative execution, and traceable evaluation, reporting gains on a 48-skill benchmark called Skill-X.
- CogniFold: Always-On Proactive Memory via Cognitive Folding
- AgentPSO: Evolving Agent Reasoning Skill via Multi-agent Particle Swarm Optimization