Chantless research trace: six months from arithmetic-MLP experiments to full-model specialists
tokenbender · x · 2026-10-14
tokenbender shares the backstory of Chantless: roughly six months of research, with early experiments limited to arithmetic operations and MLPs, now extended to specialist extraction across entire models — with more results coming soon.
Related event: Chantless: Taste-Grown Custom Models with 8B Beating Giants by 107.8%(3 posts)→
More from Research
- New paper predicts alignment generalization from value representations, builds first LLM value taxonomy — VeredShwartz · 2026-10-14
- 'Two-way RL' for task-level model compression: building wafer-thin hyper-specialized models — tokenbender · 2026-10-14
- Custom harness lifts GLM-5.3-Flash from 4.4% to 61.1% on long-context QA — omarsar0 · 2026-10-14
- Mathlib Standards Too High for AI Wave, Argues Resolution's Geoffrey Irving — geoffreyirving · 2026-10-14
- Resolution Grants $1M to Mathlib Initiative to Prepare for AI-Formalized Math — geoffreyirving · 2026-10-14
- NVIDIA's Nemotron hits gold at both IOI and IMO: 535.4/600 at IOI, 30/42 at IMO — NVIDIAAI · 2026-10-14