Recirculation Architecture Reduces Perplexity Across Gemma3 Models
xennygrimmato_ · x · 2026-08-20
Demonstrates the reduction in perplexity across the Gemma3 model series when using the Recirculation architecture. Comparisons show that the Recirculation transformer achieves lower perplexity than the Unrolled loop transformer.
Related event: DeepMind's Recirculation boosts LLM reasoning without retraining(10 posts)→
More from Research
- Five pharma giants federally train OpenFold3 on 20,000 private structures, +11 pts — jiqizhixin · 2026-10-16
- SWAN: a semantic-level text watermark that survives paraphrasing — burkov · 2026-10-16
- Interp researchers push back: 'networks are already clean' claim rests on dropout artifact — aryaman2020 · 2026-10-16
- LayerTrim: training-free pruning keeps reasoning LLMs intact on AIME — SeoulNatlUniv · 2026-10-16
- Simular presents neuro-symbolic computer use agents at BayLearn — xwang_lk · 2026-10-16
- Google's Hartmut Neven disputes chip test, outlines plan to link a brain to a quantum computer — Chris_Armstrong · 2026-10-16