A developer benchmarked two memory systems for Claude Code across 1,453 agent sessions, finding that instructing an agent to consult memory when none exists makes it measurably worse. The RE-call memo…
A developer's benchmark of Claude Code with a memory layer called RE-call found it improved task success from 36.1% to 58.3% over a static CLAUDE.md file, a 22.2 percentage point gain with a p-value o…
A developer reports that giving Claude Code a project memory system called RE-call reduced its mistake rate to zero across 40 paired tasks, compared to a 52.5% error rate without memory. The system al…
A developer's RE-call memory system, which never calls an LLM, scores lower on temporal reasoning (0.408 vs. 0.567) than Mem0 on the BEAM benchmark, but avoids any LLM cost and keeps data in Postgres.…
A developer discovered that the widely-cited LOCOMO benchmark for AI memory excludes 22.5% of its questions—those testing whether a system can correctly say 'I don't know'—and instructs models to neve…
A developer fine-tuned the all-MiniLM-L6-v2 embedder on a 14-document personal corpus and measured zero improvement in retrieval metrics (Δ+0.00 MRR, Δ+0.00 nDCG@10), concluding that off-the-shelf emb…
RE-call, a retrieval-augmented generation system, introduces a benchmark that evaluates not only ranking quality but also a system's ability to abstain when no answer exists. Testing on 14 answerable …
A developer building the RE-call memory system for AI agents introduces honesty guards to prevent retrieval from returning confident noise. The system wraps retrieval with signals like gap_warning (th…
A developer built RE-call, a retrieval-augmented memory system for AI agents that runs entirely on PostgreSQL, using pgvector for dense vector search and built-in full-text search, fused via Reciproca…