The authors treat external memory as a capacity-constrained population of reusable experiences.
Delayed recommendation feedback can guide useful evolution of agent memory
TIDE assigns noisy outcomes to referenced memories, then updates a capacity-limited memory store for better performance on future tasks.
Chinese Tech
Shanwen Mao · Mingming Li · Hao Zhang · Zhiheng Li · Yige Wang · Penghua Yu · +1 more
Harbin Institute of Technology · Alibaba Group · Institute of Automation, Chinese Academy of Sciences
Research Digest··3 min read
Mao et al.
Why this paper
From Institute of Automation, Chinese Academy of Sciences and 2 others
In one line
Delayed recommendation feedback evolves agent memory via three-stage credit assignment, achieving +7.75% MEG.
What we could check
- ·No code link found
- ·No weights link found
- ·No dataset link found
- ·No compute details found
- ✓Limitations stated by the authors
- ✓Reports numbers on named benchmarks
Observed from the paper text and links we have. Absence here means we did not find it, not that it does not exist.
§