What Accuracy Hides In Learned Transformer Programs
Transformer Programs (TP) (Friedman et al., 2023) constrain a Transformer's architecture so that its learned computation can be extracted as readable Python code.
ProofPaper ↗
Key points
- TP expose learned computations, but token accuracy can hide incorrect complete answers.
- We compared token accuracy with accuracy on complete output sequences and checked for failures missed by sampled tests.
- We examined a counting task where a program achieved 100% accuracy on sampled test sequences but failed on another input using the symbols and lengths allowed during training.
- The TP training process can lead to accuracy losses late in training.
Sources (1)
- [1]What Accuracy Hides In Learned Transformer ProgramsarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 16, 03:45 AM
Transformer Programs (TP) (Friedman et al., 2023) constrain a Transformer's architecture so that its learned computation can be extracted as readable Python code.
TP expose learned computations, but token accuracy can hide incorrect complete answers.
Extractive summary: sentences quoted from the sources.
Before this
- Oct 16, 2026Welcome Beam: a 501B open model you can own, customize, and scale
- Oct 15, 2026Granite Switch 4.1: Chat, Decisions, and Speech from One Endpoint
- Oct 15, 2026Pearl-Chat: Engineering a native Luganda language model from first principles in JAX and Flax NNX
- Oct 15, 2026Local and Distributed Large LLM Inference with MLX and RDMA on Mac Studio
- Oct 14, 2026Show HN: Building an LLM from Scratch in TypeScript
- Oct 14, 2026mistralai/Timestral-209M