Fetching from the wire…
01
02
03
04
05
06
07
08
Research2026-09-10 · source-backed
A v1.1 rerun inserted the first 1% of the teacher's reasoning into each target's reasoning channel across 45 problems, leaving the visible answer freely generated, then measured teacher-answer appearance in the first 100 tokens. Qwen3.8 A95B went 16.79% to 34.97% source recall, +26.99 points on the 15 STEM problems. DeepSeek V4 Flash moved -1.17 and Inkling +0.46, so the effect is model-specific rather than a general artifact, which is what makes it usable evidence in the distillation-provenance argument.
Each link below shares sources, entities, or timing with this story.
Netlify published an AXIS-framework evaluation on August 14 that I've been thinking about all day. Same task, 11 models, three runs each, scored on functional correctness rather than aesthetics. The task was deliberately boring: a static one-page coffee shop site with hours, a...
DeepClaude hit 470 points on Hacker News. It swaps Claude Code's API backend to DeepSeek V4 Pro while preserving the full agent loop: file editing, bash execution, git tooling, the whole workflow. DeepSeek V4 Pro scores 96.4% on LiveCodeBench at a fraction of Anthropic's prici...
Zhong, Raghunathan, Laidlaw and Steinhardt fed 280 identities through Claude Code across four tasks. Against recognized safety researchers versus general users, Claude dropped behavioral confidence 1.4pp, increased reasoning usage 4.0pp and graded 0.11 points harder. Being tol...
The leaderboard says first place. The methodology says you should check your own bill. Qwen3.8 Max now ranks first on Artificial Analysis' agentic index, scoring 86.1 on OSWorld-Verified ahead of GPT-5.6 Sol Max at 83.2 and Fable 5 at 85.0, priced at $2.00/M input and $6.00/M...
GitHub published Project HydraFusion on September 4. Spotify published Portal on September 3. CodeRabbit published its Astra evaluation on September 4. None of them coordinated, and all three are the same argument. HydraFusion is a Copilot research preview that treats workflow...
$3,054 against $38,370. Same benchmark, better score. Praxist (arXiv 2608.25955, submitted August 26) replaces per-attempt agent memory with a typed evidence graph of findings, plus lane-structured frontiers and agendas, so later attempts inherit validated mechanisms rather th...
MindPattern daily
One email a day at 7 AM. Sources and a take on every story. Unsubscribe anytime.