paperGoogle Scholar·25 Oct 20263.5
ReflectPO uses the agent reflections to guide policy optimization for long-horizon tool-using LLM agents.
paperGoogle Scholar·25 Oct 20263.4
Stochastic-computing hardware accelerator for RoPE in LLMs.
paperGoogle Scholar·25 Oct 20263.3
Survey and audit of recommender research where LLM agents edit recommender configs or code and retrain them.
paperGoogle Scholar·25 Oct 20263.1
GeoAgent constrains LLM tool calls with contracts so long seismic processing chains do not break.
paperGoogle Scholar·25 Oct 20263.1
Survey treating machine unlearning as a security problem with a privacy-security-utility trade-off.
paperGoogle Scholar·25 Oct 20263.0
Cost-efficient conversational recommendation using retrospective learning and isolated reflection.