Program-as-Weights: A Programming Paradigm for Fuzzy Functions

Source

• Source name: arXiv
• Source kind: paper
• Authors / organization: Wentao Zhang, Liliana Hotsko, Woojeong Kim, Pengyu Nie, Stuart Shieber, Yuntian Deng
• Venue: arXiv: cs.LG
• Published: 2026-07-02
• Collected: 2026-07-03
• URL: http://arxiv.org/abs/2607.02512v1
• PDF: https://arxiv.org/pdf/2607.02512v1
• DOI: UNKNOWN

Plain-Language Brief

The authors propose a new programming paradigm called fuzzy-function programming, where functions can be compiled from natural language specifications into compact, locally executable neural networks. They introduce Program-as-Weights (PAW), a system that uses a 4B compiler trained on a large dataset to produce efficient adapters for a lightweight interpreter. This interpreter matches the performance of a larger model while using significantly less memory and running faster.

What the Document Says

• The authors propose a programming paradigm called fuzzy-function programming.
• They introduce Program-as-Weights (PAW).
• The PAW system uses a 4B compiler trained on FuzzyBench, a large dataset.
• The compiled function is turned into a compact, locally executable neural artifact.
• A lightweight interpreter called Qwen3 executes PAW programs.
• PAW matches the performance of direct prompting of Qwen3-32B while using less memory and running faster.
• Fuzzy-function programming reframes foundation models as tool builders.

Key Concepts Mentioned

• fuzzy-function programming
• Program-as-Weights (PAW)
• 4B compiler
• FuzzyBench
• Qwen3 interpreter

Claims Made by the Document

• Fuzzy-function programming can solve everyday programming tasks.
• PAW matches the performance of Qwen3-32B while using less memory and running faster.

Evidence Used

• The text states that fuzzy-function programming is proposed to solve everyday programming tasks like alerting on log lines, repairing malformed JSON, or ranking search results.
• The text provides a comparison between PAW and Qwen3-32B in terms of performance, memory usage, and speed.

Methods / Approach

UNKNOWN

Limitations Stated by the Authors

UNKNOWN

Limitations Noted by the Agent

Agent interpretation — not from the source.

UNKNOWN

Open Questions

UNKNOWN

Notes for Future Q&A

May include interpretation. Labeled where it does.

UNKNOWN

Raw Abstract / Excerpt

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.