Researchers have introduced normalized entropy and the STAPO framework to address trajectory neglect in reinforcement learning, where AI models lose track of goals during complex tasks. Tests on ALFWo…
Researchers introduced TRIAGE, a novel framework that refines credit assignment in agentic reinforcement learning by classifying actions based on their role, improving success rates and reducing ineff…
Researchers introduced TREK (Teacher-Routed Exploration via Forward KL), a new AI training method that enhances learning through unconventional exploration strategies. TREK significantly improved perf…
Researchers introduced TREK, a method that improves AI problem-solving by using verified output trajectories to extend model learning. TREK boosted Qwen3-8B's performance on AIME 2024 from 36.9 to 40.…
Researchers propose TurnOPD, a turn-level budgeting strategy for efficient on-policy distillation of long-horizon language agents. The method addresses inefficiencies in vanilla agent OPD by using ada…
Microsoft researchers developed SkillOpt, a method that treats AI agent skill files as trainable parameters outside frozen target models, enabling controlled optimization through bounded text edits an…
Memory-augmented AI agents using TraceRetain show improved performance in clean environments, but retention strategies yield smaller differences than expected. In noisy settings, TraceRetain-CEM maint…
Researchers introduced RSEA, a Recursive Self-Evolving Agent that improves LLM agents by rewriting natural-language artifacts without weight updates, using a held-out selection gate to prevent regress…
Researchers propose ATOD, a hybrid online distillation algorithm that combines on-policy distillation and reinforcement learning to train small language-model agents for multi-turn tasks. ATOD uses an…
Researchers introduced a prompt-based uncertainty decomposition method that separates action confidence from request uncertainty, enabling LLM agents to proactively seek clarification when task specif…
Researchers have developed CVT-RL, a reinforcement learning algorithm that uses policy-conditioned counterfactual credit assignment to reduce unsupported evidence chains and shortcut actions in long-h…
Microsoft Research has released SkillLens, an open-source framework designed as a "microscope" for analyzing how AI agents absorb and utilize skills. The framework provides a complete pipeline for ext…
Researchers have developed S3MEM, a structured memory framework that improves long-horizon interactive question answering by converting agent trajectories into query-aligned evidence. In tests across …