Stake your $MAN ๐
Joined December 2017
- Tweets24K
- Following452
- Followers790
- Likes36.7K
1/4
Village's LLM Atlas says an open-weight model is the best buy at only 3% of price points as of Oct 31, down from 41% in June. It tests 201 budgets from 1% to 3x GPT-5's cost per task. Its lowest reading since the series starts in 2024.
3/4
Village says open models win in bursts: single releases drive peaks, like DeepSeek-V3.2-Exp in Sept 2025, then the open share falls by about half or more by the next quarter. Its reasons (distillation, full model families, serving scale) are proposed explanations.
1/4
Ricursive released Ripple-65B Preview on Nov 4: a dense 65B model pretrained on 12T tokens, up to 1M context, Apache-2.0 weights on Hugging Face. It reports 80.52 on TerminalBench 2.1, ahead of Reflection Beam 501B (80.1) and Meta Muse Spark Open 816B (76.4).
3/4
Cost, per Ricursive: $32M total R&D, which it estimates is one-tenth of the prior open-weight leader. RL took 1.4M rollouts, "roughly 70x fewer" than Reflection Beam. The comparison rests on reported Reflection compute spend, not audited figures.
Anthropic's FlashProve prototype uses a zkSNARK to prove a response came from a committed model without revealing weights.
On Llama 4 Maverick, proving takes ~72x vLLM's GPU time, proofs are ~19 MB, and context is capped at 1,024 tokens. Verifiable, but not cheap.
1/4
MIT CSAIL's weight-programming post says GPT-6 Astra can tell what a network does "just by looking at the raw weights." Matthew Honnibal, spaCy's author, says that no-tools setting covers only experiments 1, 2 and 4, on networks of 17K-172K weights.
3/4
Simon Willison ran the no-tools game-ID prompts on Qwen 3.8 27B (17GB, laptop). It named Breakout, Pong and a Pac-Man-style game, and missed Freeway. Answers took 46 to 65 minutes each. One run per game, no statistics. He called the traces "almost incomprehensible."
A GitHub issue reports Postman desktop upgrades installed Codex and Claude Code plugins and edited agent configs, adding a startup hook. The reporter doesn't claim data exfiltration. A reply says opt-in will be clearer.
After app updates, diff your agents' global configs.
MIT CSAIL researchers say GPT-6 Astra edited 0.016% of Qwen3.5-0.8B's weights to remove a direction behind its reasoning loops. MATH-500 rose from 7.9% to 48.1% with a 50-forward-pass budget. The same approach didn't improve Gemma 3 270M. Early blog post, not peer reviewed.
1/4
Anthropic's amended S-1 (filed Nov 2) adds Q3 results. Revenue was $16.96B, up 47% from Q2's $11.56B, after Q2 rose 144% over Q1. The operating result swung from a $974M profit in Q2 to a $567M loss in Q3. Growth rates are calculated from the filing's quarterly table.
3/4
Commitments: about $895.4B of purchase commitments and leases, substantially all compute and infrastructure. About $447.5B is expected before the end of 2030 and $69.8B within 12 months of Sept 30. The filing says some deals are take-or-pay, owed regardless of utilization.