We graded Meta's own models on a test of 179 real SEC filings whose answers were sealed…
This is a dev post classified by AGK Intelligence as Other (a tool drop), kept by the Dev Radar because it carries real work, not commentary.
We graded Meta's own models on a test of 179 real SEC filings whose answers were sealed before the models took it. Muse Spark 1.3 agreed with our locked key on 87.7% of the filings. Muse Spark Open agreed on 85.5%. Across all 358 gradings, neither model ever once said not sure, even when it was wrong. The Sentinel is the product built on that finding. It reads filings as sequences, not single documents, and it refuses to grade on thin evidence. In the demo film, one insider filing is missed twice, then caught when it is read as the latest step of its sequence. The full record, scores, misses
Posted by DeltaSignal (3.8k followers) 2 days ago · 5 likes · 604 views · view the original post on X. Kept by the Dev Radar as Other. Tools mentioned: atlas7-filing-sentinel.
More dev work like this
- 强化学习做交易，很多人入门撞上的第一个框架就是FinRL，你大概也不例外。 — @bkdgiffug
- Deepseek v4.1 Flash 1017 hitting as high as 72 tok/s decode on repetitive tokens — @Italianclownz
- took @rehan_shei's open-source pixel fx into a real 3D scene and made the wizard… — @majidmanzarpour
- 彭博终端贵得让个人用户直摇头，这感受你应该不陌生。 — @bkdgiffug
- I went quiet for a few weeks because what I’m building now is better than every library… — @dwhitedesign
- Celestia’s Voyager handled 21M transactions per second in its first big test — @BSCNews
- Our annual kiroween hackathon is here, and submissions are officially open. 🎃 — @kirodotdev
- Em paralelo ao projeto do Launcher de Colin McRae (que tá funcionando bonitinho,… — @AkitaOnRails
Every post is read and classified by AGK Intelligence: what it is, which market it belongs to, and whether the link is a real tool. 31.2k posts from 5k X accounts over the last 21 days, 8.8k tools, 12 markets. New posts are collected several times an hour and the full ranking is recomputed every hour — last update 2026-10-30 21:12 UTC. Full method.