7 accepted papers
Traditional value models $V^{\pi}$ in LLM reinforcement learning face a coupling dilemma: they require synchronous training alongside the updating policy $\pi$, causing inefficiency and overfitting. In this paper, we propose $V_0$, a generalist value model that decouples value estimation from specif…
As LLM-based agents are increasingly deployed in real-world workflows, existing agent benchmarks---often built on idealized, noise-free assumptions---fall short of characterizing agents' robustness under imperfect user instructions and unreliable tool feedback. To address this gap, we introduce **Ag…
Recent large language models (LLMs) achieve strong performance in generating promising reasoning paths for complex tasks. However, despite powerful generation ability, LLMs remain weak at verifying their own answers, revealing a persistent capability asymmetry between generation and self-verificatio…
Large language models face challenges in long-context question answering, where key evidence of a query may be dispersed across millions of tokens. Existing works equip large language models with a memory corpus that is dynamically updated during a single-pass document scan, also known as the "memor…
Long-horizon agentic reasoning necessitates effectively compressing growing interaction histories into a limited context window. Most existing memory systems serialize history as text, where token-level cost is uniform and scales linearly with length, often spending scarce budget on low-value detail…
Equipping agents with interactive environments and verifiable tasks for self-exploration is essential for cultivating generalist agents capable of adapting to diverse scenarios. However, high-quality agentic data remain critically scarce, and existing synthesis methods suffer from significant limita…
As LLMs with agentic abilities are increasingly deployed in real-life scenarios, existing benchmarks fail to capture their inherent complexity of handling extensive information, leveraging diverse resources, and managing dynamic user interactions. To address this gap, we introduce VitaBench, a chall…