{"slug":"remember-memory","label":"Remember Memory","item_count":3,"day_count":3,"source_count":2,"first_seen":"2026-11-03T21:21:14+00:00","last_updated":"2026-11-19T16:29:16+00:00","generated_at":"2026-11-21T00:13:05.218527+00:00","sources":["arxiv_cs_ai","arxiv_llm_reliability"],"days":[{"date":"2026-11-03","items":[{"title":"Route to Remember: Adaptive Memory Selection for LLM Agents","url":"http://arxiv.org/abs/2611.05636v1","source":"arxiv_llm_reliability","type":"paper","summary_1line":"The memory of production agents is increasingly becoming a bottleneck and is built almost exclusively from text: key points are extracted from out-of-limit context and retrieved into context at query time. Therefore,...","why_it_matters":"Matches feed focus: agent, harness, evaluation.","sid":"043490c38692d4af","published":"2026-11-03T21:21:14+00:00","editor_note":"Sets the starting problem: text memory extracted from overflow context is a retrieval bottleneck."}]},{"date":"2026-11-12","items":[{"title":"We Are What We Remember: Mental-State-Driven Memory for Long-Term Character Dialogue","url":"http://arxiv.org/abs/2611.16223v1","source":"arxiv_llm_reliability","type":"paper","summary_1line":"Shared history in human dialogue is not merely a repository of facts, but the foundation of evolving relationships and mutual understanding. However, mainstream conversational memory systems rely on passive semantic s...","why_it_matters":"Matches feed focus: agent, eval.","sid":"f48f97340f1b60b5","published":"2026-11-12T03:17:19+00:00","editor_note":"Extends the theme from task agents to conversational memory that tracks evolving relationships."}]},{"date":"2026-11-19","items":[{"title":"How Much Should an Agent Remember? Deployment Lessons from Self-Distilled Memory for LLM Agents","url":"http://arxiv.org/abs/2611.26361v1","source":"arxiv_cs_ai","type":"paper","summary_1line":"LLM agents on long-horizon tasks fail in recurring, procedural ways. They mis-paginate an API, resolve the wrong entity, or answer when no answer was asked. We built a pipeline that lets an agent learn from its own hi...","why_it_matters":"Matches feed focus: agent.","sid":"32373812989fabcd","published":"2026-11-19T16:29:16+00:00","editor_note":"Adds deployment evidence: memory learned from the agent's own history, aimed at named procedural failure modes."}]}],"editorial":{"tldr":"This is a loose grouping of three independent arXiv papers on LLM agent memory, linked by topic rather than one event. It opened on Nov 3 with a paper on choosing which text memories to retrieve, then moved to memory for long-running character dialogue on Nov 12.","stale":false,"whats_new":"A Nov 19 paper reports deployment lessons from agents that distill memory from their own history to avoid repeat procedural failures such as mis-paginating an API or resolving the wrong entity.","why_it_matters":"Agent memory is shifting from one fixed retrieval step to a design choice about what to store and when to inject it, which affects context cost and repeat-failure rates.","take_for_builders":"Log your agent's recurring failures (pagination, entity resolution, unprompted answers) before adding a memory layer, so you can measure whether stored lessons actually reduce them.","beats":[{"kicker":"RETRIEVAL","tone":"launch","headline":"Adaptive selection of extracted text memories","summary":"Frames production agent memory as a bottleneck built from text key points pulled out of over-limit context.","sids":["043490c38692d4af"]},{"kicker":"DIALOGUE","tone":"rising","headline":"Mental-state-driven memory for long-term character dialogue","summary":"Argues passive semantic retrieval misses how shared history shapes a relationship.","sids":["f48f97340f1b60b5"]},{"kicker":"DEPLOYMENT","tone":"now","headline":"Self-distilled memory targets recurring procedural failures","summary":"Agents learn from their own trajectories, and the paper asks how much they should keep.","sids":["32373812989fabcd"]}],"open_questions":["Does self-distilled memory reduce repeat failures on public long-horizon agent benchmarks, not just the authors' tasks?","At what memory size does retrieval start hurting answers instead of helping them?","Will any agent framework ship adaptive memory selection as a built-in option?"],"generated_at":"2026-11-21T00:12:20Z"}}