Jun 30, 2026 · Artificial Intelligence

Why One Extra Loop Is All a 7B Model Needs – LoopCoder‑v2’s Surprising Sweet Spot

LoopCoder‑v2, a 7B LLM, gains a massive boost on SWE‑bench Verified (43.0 → 64.4) by adding just one test‑time loop, while additional loops cause performance to collapse, a finding explained through detailed probe analysis of hidden‑state convergence, attention re‑routing, and a constant “position‑mismatch tax”.

AI model efficiencyLLM loopingLoopCoder-v2

0 likes · 8 min read

Why One Extra Loop Is All a 7B Model Needs – LoopCoder‑v2’s Surprising Sweet Spot

Wuming AI

Apr 19, 2026 · Artificial Intelligence

Why Bigger LLMs Aren’t Smarter: Karpathy Blames Junk Training Data

Karpathy argues that the rapid growth of large language models is driven more by noisy, low‑quality training data than by a need for greater intelligence, urging a split between clean cognition cores and external memory to achieve smarter, more efficient AI.

AI model efficiencyKarpathyLLM scaling

0 likes · 5 min read

Why Bigger LLMs Aren’t Smarter: Karpathy Blames Junk Training Data