Sentence Transformers v6.3 adds token pooling-aware training for cheaper multi-vector retrieval
CShorten30 · x · 2026-11-11
Sentence Transformers v6.3 ships token pooling-aware training, teaching multi-vector models to preserve retrieval quality with fewer document vectors. The hpool paper's author notes the progression: pooling was already a free lunch, proper training made it cheaper than free, and the new kernel makes it free AND fast.
More from coding & agent
- Perplexity red-team finds 9 of 12 AI agents barely probe the VM security boundary — perplexity_ai · 2026-11-12
- Perplexity research: standard CTF red-teaming steers agents away from the security boundary — perplexity_ai · 2026-11-12
- ChatGPT Business plans now work in Warp, Notion, v0 and more via Sign in with ChatGPT — OpenAIDevs · 2026-11-12
- Cloudflare's weekend-trained hybrid model cuts agent cost from 40K to 1.8K tokens — TheTuringPost · 2026-11-12
- Ex-Meta CV researcher builds AdaL v2, a chat-driven video agent harness for GTM — Scobleizer · 2026-11-12
- chrisalbon: Show me 3 months and 1000 prompts, not one magic prompt — chrisalbon · 2026-11-12