DSPy and GEPA users are writing custom proposers to avoid overfitting
dbreunig · x · 2026-07-24
The post discusses how people using DSPy/GEPA often write custom proposers to reduce overfitting and better tailor proposer instructions to their tasks.
The author says they are trying out a few ideas to make that easier and asks for issues, comments, or PRs so the space can stay focused on proposer-concept iteration.
Related event: DSPy and GEPA Users Adopt Custom Proposers to Prevent Overfitting(2 posts)→
More from coding & agent
- Open-Source AI Hedge Fund Project Adds Backtesting and Benchmarking — virattt · 2026-07-24
- AI Hedge Fund Update: Full Backtesting with Point-in-Time Data — virattt · 2026-07-24
- Building Claude AI SDR Agents: The Challenge of Confidence Scoring in Multi-Source Data — dettol99perc · 2026-07-24
- Optimizing Local AI Agents: Fixing Multi-Source Search Aggregation and Context Bleed — Glad-Finance4354 · 2026-07-24
- Hermes Agent Launches Open-Source Plugin to Turn Messy PDFs into Structured Data — bigaiguy · 2026-07-24
- Configure an MCP in 60s to Turn Claude Into an All-in-One Video Editing Suite — dr_cintas · 2026-07-24