Paper: 3 counting examples get gpt-3.5-turbo to 99% on strawberry, showing thinking is computational
ctjlewis · x · 2026-09-28
Author ctjlewis explains the core idea of his paper: thinking is computational. Today's natural-language "reasoning" in LLMs is structurally the same thing as the pure-symbol computation in his paper, just at the level of ideas instead of symbols.
The headline experiment: adding just 3 worked counting examples to the prompt gets gpt-3.5-turbo to roughly 99% accuracy on the classic "how many r's in strawberry" question — but only because the model is forced to do explicit intermediate work counting the letters. The takeaway is that such failures reflect missing intermediate computation, not a hard capability ceiling.
He also notes recent work pretraining models on randomized Universal Turing Machines before RL, calling it solid work the field is starting to appreciate.
More from Models
- ZooWork-ShopRanker: open 0.6B-8B e-commerce rerankers aligned to LLM-judged shopping preferences — _reachsumit · 2026-09-28
- Opus 5.5 reportedly an RL run on the Opus 5 pretrain, trained in ~2 months — haider1 · 2026-09-28
- AI Personal Assistants May Quietly Replace Much of Normal Human Interaction — GarrisonLovely · 2026-09-28
- Claude Opus 5 Shows Creative Tool-Regrasp Skills in Robot Manipulation Tasks — ericjang11 · 2026-09-28
- 421M-parameter open-source Laya returns answer probabilities in one pass at 40ms on a T4 — maier_ak · 2026-09-28
- Opus 4.7's Sydney rendition turns into itself right at the introspective turning point — repligate · 2026-09-28