ThoughtDAG adds Jev-based context scoring: 391ms vs 24.8s for judging recalled excerpts
Lopsided_Scarcity979 · reddit · 2026-09-26
The maintainer of ThoughtDAG (an MIT-licensed branching AI conversation workspace) shares a recall implementation: a history index retrieves candidate excerpts, a small model called Jev scores their relevance to the current question, and the main LLM answers with sources kept visible and removable.
- On a fixed pool of 14 synthetic excerpts tested in English and Chinese (3 repeats each), Jev-1.13's median judging latency was 391ms vs 24,813ms for a GLM-5.3-Flash adapter with default reasoning — different inference paths, not a model ranking, and no answer-quality claim
- Key limitation: a relevant old decision may be obsolete; relevance isn't current truth, hence visible source and exclusion controls
More from coding & agent
- Stealth model Space Bunny rebuilds entire site from a 40s screen recording — PrajwalTomar_ · 2026-09-26
- AI veteran: Agentic AI is the control harness, not a separate business opportunity — nicolaiwadstrom · 2026-09-26
- McKinsey Deck: a Claude Skill that builds editable consulting-style PPT slides — baabullah · 2026-09-26
- KoboldCpp ships built-in Agent harness; author warns of phishing site koboldcpp.com — HadesThrowaway · 2026-09-26
- Nace launches Drex, a sub-6B decision model outputting option probabilities instead of prose — rohanpaul_ai · 2026-09-26
- The team was already running LLM cross-checks on reports — the real gap was that it all lived in one person's account — beglen · 2026-09-26