https://lens.lumihelia.com/en/papers/ Grounded readings, every point traceable to the source. en https://lens.lumihelia.com/en/papers/humans-almanac.html https://lens.lumihelia.com/en/papers/humans-almanac.html Sat, 06 Jun 2026 00:00:00 +0000 This paper introduces ALMANAC, the first dataset of Action-Level Mental Model Annotations for human-agent collaboration. Although Large Language Model (LLM) agents possess multi-step reasoning and pla… Agent Collaboration Mental Models ALMANAC Map Task Behavior Prediction https://lens.lumihelia.com/en/papers/agent-memory.html https://lens.lumihelia.com/en/papers/agent-memory.html Sat, 06 Jun 2026 00:00:00 +0000 This paper presents the first system-level characterization of Agent Memory Systems for Large Language Models (LLMs). As LLM agents are increasingly deployed in long-horizon tasks requiring continuous… Agent Memory System-level Characterization Prefill Overhead Memory Construction Freshness and Latency Energy Overhead https://lens.lumihelia.com/en/papers/the-era-of-experience.html https://lens.lumihelia.com/en/papers/the-era-of-experience.html Sat, 06 Jun 2026 00:00:00 +0000 This paper is written by David Silver and Richard S. Sutton (the text is a preprint of a chapter in the forthcoming book Designing an Intelligence to be published by MIT Press). The paper explores how… experience streams grounded rewards bi-level optimisation self-play world models