MemRole is a framework that explicitly models the memory-to-decision process for large language model (LLM) role‑playing agents, enabling them to make role‑consistent decisions over long narrative horizons. It introduces Long‑Horizon Memory Thinking (LHMT), a reasoning schema that decomposes hidden reasoning into context understanding, memory activation, and response strategy planning, and is trained on a novel dataset that limits each character’s memory pool to what it could plausibly know at the current narrative point. The framework also includes Self‑Refining On‑Policy Self‑Distillation (SR‑OPSD), where the model critiques its own rollouts and distills token‑level guidance back into the policy, resulting in a 15.2‑point improvement over its backbone and competitive performance with frontier LLMs such as GPT‑5.4.
arXiv:2607. 01224v1 Announce Type: new Abstract: Memory expertise is a learned skill: knowing what to encode, when to retrieve, and how to organize knowledge--a capacity known in cognitive science as metamemory.
arXiv:2608. 10108v1 Announce Type: new Abstract: Long-horizon agents accumulate trajectories spanning hundreds of interleaved reasoning, action, and observation steps, where answering a query may depend on evidence buried far back in the history.
M-Plan is a closed‑loop multi‑agent system that combines long‑term and short‑term memory to improve large language model agents on long‑horizon, multi‑turn tasks. It uses task‑specific milestones and local hints to generate global‑local guidance, forming short‑term memory during execution, and then optimizes and stores relevant information in long‑term memory for future use. Experiments on ALFWorld, Webshop, and PDDL show that M‑Plan outperforms baseline methods in both accuracy and efficiency.
arXiv:2608. 12626v1 Announce Type: cross Abstract: Strategic reasoning in Large Language Models (LLMs) within long-horizon environments is often limited by inconsistent subgoals.
arXiv:2607. 01480v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards (RLVR), along with recent selfdistillation variants such as SDPO, evaluates each rollout against a verifier and updates the policy from that episode-level signal.
Large Language Model (LLM) agents increasingly rely on external memory systems to accumulate experience across tasks. Yet nearly all existing approaches, from graph-structured memories to reflective insight stores, access memory through fixed, hand-designed heuristics.
arXiv:2606. 24064v1 Announce Type: new Abstract: Distilling reasoning capabilities from strong to weak language models typically involves imitating specific solution trajectories, effectively transferring what to answer rather than how to reason.
The paper introduces MemCryst, a framework that enables lifelong learning for large language model agents by jointly adapting memory and skills while keeping the language model frozen. MemCryst stores recurring episodes in memory to support skill creation and revision, and uses the resulting skills to guide future interactions, which in turn generate new memory. Experiments across ten diverse settings—including embodied control, long-document reading, conversation, mathematics, and code—show that MemCryst outperforms memory-only and skill-only baselines, achieving 63.17% success on ALFWorld.
Autonomous computer-use agents are increasingly applied to long-horizon tasks requiring coordinated application calls, persistent state tracking, and verifier-sensitive writes, yet they remain prone t...
The paper investigates whether reinforcement learning (RL) truly endows large language models (LLMs) with reasoning skills or merely tunes them to exploit prompt information. By training small LLMs on the finite logical domain of Tic‑Tac‑Toe and applying Group Relative Policy Optimization (GRPO), the authors show that RL improves accuracy and robustness. However, mechanistic interpretability using sparse autoencoders reveals that the models mainly rely on explicit prompt cues—such as turn order, game progression, and board occupancy—while higher‑level concepts like board symmetries and strategic forks remain weakly represented, suggesting that perceived reasoning may be an artifact of task‑specific adaptations.
arXiv:2607. 13591v1 Announce Type: cross Abstract: Large Language Model (LLM) agents increasingly rely on external memory systems to accumulate experience across tasks.
arXiv:2507. 05257v4 Announce Type: replace-cross Abstract: Recent benchmarks for Large Language Model (LLM) agents primarily focus on evaluating reasoning, planning, and execution capabilities, while another critical component-memory, encompassing how agents memorize, update, and retrieve long-term information-is under-evaluated due to the lack of benchmarks.