TokenWorth: Nanjing Univ. Teaches Lifelong Agents Which Memories Are Worth Their Token Cost
nanjinguniv · hf · 2026-10-20
Nanjing University proposes TokenWorth, tackling which past experiences lifelong agents should retain and reuse under tight generation, interaction, and memory budgets.
- Finding: under resource scarcity, sophisticated lifelong-memory methods can offer little advantage over simple retrieval or memory-free agents—relevant but ineffective memories consume scarce context and hurt downstream execution
- Shifts the problem from retrieving relevant memories to learning memory worth, where every memory decision carries higher stakes
- A lightweight online framework improves lifelong agent success while reducing token usage
More from coding & agent
- Indie dev says AI removed the model bottleneck: solo-building a physics excavator sim — Michael_J_Black · 2026-10-20
- HSPR Recovers Full Agent Benchmark Scores From Small Subsets via Hierarchical Posterior Inference — SUDA · 2026-10-20
- AdaExploreBench Finds LLM Agents Struggle at Proactive Exploration in Evolving Environments — HKUST · 2026-10-20
- Our OAuth discovery doc and Python docs give agents opposite signup instructions — bshivarthy · 2026-10-20
- Fo launches MCP to let AI agents control and sense physical devices, 100+ devs already building — heyshrutimishra · 2026-10-20
- Open-source humanoid Asimov 1 ships docs: assembly guides, sim-to-real training, browser demo — Sentdex · 2026-10-20