Program-as-Weights: A Programming Paradigm for Fuzzy Functions
For developers needing to implement fuzzy functions (e.g., log alerting, JSON repair) without relying on expensive LLM APIs, PAW offers a locally executable, reproducible, and cost-effective alternative.
The authors propose fuzzy-function programming, where a natural-language specification is compiled into a compact neural artifact. Their Program-as-Weights (PAW) approach, using a 4B compiler and 0.6B interpreter, matches the performance of direct Qwen3-32B prompting while using 1/50th the inference memory and running at 30 tokens/s on a MacBook M3.
Many everyday programming tasks resist clean rule-based implementation, such as alerting on important log lines, repairing malformed JSON, or ranking search results by intent, and are increasingly outsourced to large language model APIs at the cost of locality, reproducibility, and price. We propose fuzzy-function programming: compiling such a function from a natural-language specification into a compact, locally-executable neural artifact. We instantiate this paradigm with Program-as-Weights (PAW), in which a 4B compiler trained on FuzzyBench, a 10M-example dataset we release, emits parameter-efficient adapters for a frozen, lightweight interpreter. A 0.6B Qwen3 interpreter executing PAW programs matches the performance of direct prompting of Qwen3-32B, while using roughly one fiftieth of the inference memory and running at 30 tokens/s on a MacBook M3. PAW reframes the foundation model from a per-input problem solver into a tool builder: invoked once per function definition, it produces a small reusable artifact whose subsequent calls per function application are cheap and offline.