Ampersand 0.1 Nearly Ties Fable 5.1 on Collaboration Eval, Starting from GLM 5.3
gharik · x · 2026-10-21
humans& published eval results for Ampersand 0.1 on its "Goal Progress minus Interaction Costs" benchmark with an adjustable interaction-cost weight.
- Ampersand 0.1 scores 12.94, nearly matching top-ranked Fable 5.1 at 12.95, ahead of Gemini 3.8 Flash (12.92), Muse Spark 1.3, Kimi K3, and Opus 5.5.
- Its total goal progress of 13.82 minus interaction costs yields the score, implying unusually low collaboration friction.
- The chart marks GLM 5.3 (9.89) as Ampersand's starting point, meaning the team post-trained it into a top collaboration model.
More from Models
- Ampersand debuts first model '&', built on long-horizon simulation stack — niloofar_mire · 2026-10-21
- Hugging Face quietly added model file size display, a big QoL win — vboykis · 2026-10-21
- Humans& launches Ampersand, first model with privacy built in, plus a funny RL-plot joke — niloofar_mire · 2026-10-21
- A single GLM neuron keeps firing on both sci-fi self-replicating malware and real cyber risk — burny_tech · 2026-10-21
- Gemini 4 Argon tops Isoquant's inference benchmark with 4.38x mean GPU speedup — ycombinator · 2026-10-21
- Gemini 4 Argon Weekly Quota Hits 0% Without a Single Message Sent in Antigravity — koltregaskes · 2026-10-21