arXiv:2605. 24828v2 Announce Type: replace Abstract: With the continuous advancement of Large Language Models (LLMs), intelligent agents are becoming increasingly vital.
While Large Language Models (LLMs) excel as static solvers, transforming them into autonomous agents remains challenging. This transition requires continuous environmental interaction, yet current agents lack the necessary persistent procedural memory.
arXiv:2606. 29824v1 Announce Type: cross Abstract: While Large Language Models (LLMs) excel as static solvers, transforming them into autonomous agents remains challenging.
arXiv:2606. 04815v1 Announce Type: cross Abstract: Lifelong learning is essential for Large Language Model (LLM) agents operating in dynamic, interactive environments.
Multi-modal Large Language Models (MLLMs) have achieved remarkable progress in video temporal grounding with reinforcement learning for generating reasoning paths. However, existing models often produce superficial reasoning, which offers limited guidance for precise temporal localization.
MIRAGE is a new inference-time framework that enhances large language models by using a Selector to choose effective conceptual perspectives and a Reasoner to solve tasks step-by-step, aggregating multiple perspectives when needed. It is inspired by human cognitive flexibility and is designed to improve performance on complex mathematical, scientific, and logical problems. Experiments on GSM8K, MATH500, MMLU-Pro, and Game-of-24 show that MIRAGE outperforms Chain-of-Thought and diverse prompting ensembles, boosting accuracy with minimal inference overhead.
arXiv:2601. 20379v2 Announce Type: replace Abstract: Large language models (LLMs) struggle with complex, long-horizon reasoning due to instability caused by their frozen policy assumption.
The paper introduces T2T (Thickening-to-Thinning), a dynamic reward framework for large language models that mimics human learning by separating exploration and consolidation phases. During incorrect attempts, T2T encourages exploration to broaden the search space, while after correct solutions it applies length penalties to promote concise reasoning. Experiments on mathematical benchmarks across five mainstream LLMs show that T2T outperforms standard GRPO and recent baselines, improving overall reasoning performance.
arXiv:2601. 03555v3 Announce Type: replace Abstract: Training reliable tool-augmented agents remains a significant challenge, largely due to the difficulty of credit assignment in multi-step reasoning.
arXiv:2601. 21754v3 Announce Type: replace Abstract: While Large Language Models (LLMs) excel in language-based agentic tasks, their applicability to unseen, nonlinguistic environments (e.
Large language models (LLMs) are increasingly extended into deep search agents that solve complex questions through multi-step interaction with external search and browsing tools. However, existing agents often incur substantial computational and interaction costs, generating lengthy trajectories that contain redundant queries, inefficient exploration, and irrelevant observations.
APEx is a hierarchical framework that organizes a deep research agent’s interaction history into instance-level trajectory memories and category-level procedural skills. It couples these through an Executor, Distiller, and Planner, trained with a three-stage alternating GRPO paradigm to enable reward-guided skill distillation. At test time, distilled skills act as procedural priors for online Planner adaptation via skill-guided reinforcement learning, achieving state‑of‑the‑art results on seven benchmarks, outperforming GPT‑5.4 by 14.7 points and the best memory‑augmented baseline by 3.0 points.
arXiv:2606. 27136v1 Announce Type: new Abstract: For LLM agents in multi-step interactive environments, a key challenge is to make effective use of accumulated interaction experience.