Transformers Are Inherently Succinct (2025)
bearseascape
45 points
6 comments
May 04, 2026
Related Discussions
Found 5 related stories in 803.5ms across 14,015 title embeddings via pgvector HNSW
- Transformers are inherently succinct brandonb · 110 pts · June 05, 2026 · 87% similar
- Executing programs inside transformers with exponentially faster inference u1hcw9nx · 17 pts · March 12, 2026 · 55% similar
- The Trouble with Transformers rob_lh · 11 pts · April 19, 2026 · 55% similar
- Do transformers need three projections? Systematic study of QKV variants Anon84 · 137 pts · June 04, 2026 · 54% similar
- Learning Pseudorandom Numbers with Transformers pizza · 11 pts · May 03, 2026 · 54% similar
Discussion Highlights (2 comments)
pstuart
It makes sense that flowery language is more decorative than functional, but I wonder how much nuance can help shape reckoning, reasoning, and rendering -- if at all. Maybe RFC terms are all that's needed: https://datatracker.ietf.org/doc/html/rfc2119
cadamsdotcom
Seems intuitively sound; a larger model would have the ability to differentiate among a larger variety of concepts, which translates to a larger vocabulary and greater ability to use expressive tools such as imagery, metaphor etc etc. I could go on, but brevity is virtuous.