FlashProve README

项目的 README,来自 github.com/anthropics/FlashProve。合并为一个 markdown 文件:README.md。

FlashProve

Version 0.16.0 · A research artifact: not maintained, not accepting contributions. · Apache 2.0

What it is

FlashProve lets an AI provider prove which model answered a request, without revealing the model's weights.

What model it runs

FlashProve runs Llama 4 Maverick at full size, with a 1,024-token context window, on a single NVIDIA H100. The prover reaches about 110 tokens/second of prefill at 24 requests per batch (Performance).

What is proved

The protocol's soundness is proved in Lean 4. A provider that tries to pass off output that is not the committed model's is caught, except with probability about q/2^124 + 2^-101 after q hash queries. lean/THEOREM.md has the theorem and how to check it.

Built to be run by an agent

Building FlashProve, running it and checking its output take long builds and many exact comparisons. That work is meant for a coding agent. Point yours at this repository with:

Read README.md in the FlashProve repository, starting at "For agents", and AGENTS.md. Working from
the repository root as an ordinary (non-root) user, build the verifier, check the stored example
proof, and report the exit code and last line. Then tell me which other routes (toy proofs on CPU,
small model on an H100, full size) this machine can run, and ask me before starting any of them.

For agents

Run everything from the repository root as an ordinary user, never as root or under sudo. Only installing system packages needs root, so ask your user to do that. If root is the only account, create one first (how). Also read AGENTS.md for the rules on reporting results.

Hardware and software

Route What it does Needs Time
1. Verify a stored proof Builds cp_verify, checks an example proof Linux, bash, g++ 13+; under 2 GB RAM build 1–2 min, check < 2 s
2. Toy proofs Self-test of the proof library on CPU Route 1, plus make and python3; under 1.5 GB RAM ~1 min
3. Small model Builds the prover, commits, proves and verifies an 8-layer synthetic model GPU stack (below); 10 GB RAM build ~29 min, run 23–30 s
3, full size Same at Llama 4 Maverick's shape, synthetic weights, 1,001 tokens GPU stack; 280 GB RAM ~42 min
4. Llama 4 Maverick Downloads a public 4-bit conversion (232 GB) and proves one token GPU stack, curl, Python 3.13; 330 GB RAM, 245 GB disk; a calibration set that you make first hours

GPU stack (routes 3 and 4) is the only supported platform for the prover:

On minimal Ubuntu 24.04, route 1 and 2 dependencies are (needs root):

apt-get update && apt-get install -y --no-install-recommends g++ make python3 unzip

1. Build the verifier and check a stored proof

verifier/build.sh

It is silent for 1–2 minutes, then prints a line starting THE VERIFIER IS BUILT: and exits 0. A line starting build.sh: note: is a warning: quote it in your report. If your tool calls time out, run the build in the background:

echo "build started" >> verifier-build.log; nohup bash -c 'verifier/build.sh; echo "exit code $?"' >> verifier-build.log 2>&1 &

Then poll tail -n 2 verifier-build.log until the last line starts exit code. Don't start a second build meanwhile.

bin/cp_verify examples/accepted-proof/config examples/accepted-proof/ivk examples/accepted-proof/example.req examples/accepted-proof/example.y examples/accepted-proof/example.proof -id "$(cat examples/accepted-proof/ID.txt)"

Expect last line YES, exit code 0. This example is a small model with synthetic weights, so it shows only that the verifier works, not anything about a served model.

2. Toy proofs (CPU)

W=$(mktemp -d) && LOOK="$W/look" && echo "LOOK=$LOOK"
blocks/first_look.sh "${LOOK:?first run the line with mktemp}"

Expect exit code 0, and the third line from the end starts AS EXPECTED: the honest copy was accepted.

3. Prove on a GPU

examples/quickstart-gpu.sh

This builds cp_verify, cp_commit and cp_prove, then commits, proves one 64-token request and verifies it.

Full size, after that build:

W=$(mktemp -d) && WORK="$W/example" && echo "WORK=$WORK"
prover/examples/run_target_example.sh target_full_f2 "${WORK:?first run the line with mktemp}"

Expect status 0, a proof of work: line, then YES. Progress goes to $WORK/cp_commit.log. With too little memory it exits with code 2 at the start.

For the real Maverick weights, see route 4 and examples/maverick/run.sh. To commit to and serve a model of your own, see Serving and the provider's duties. Those pages are required reading before using weights that must stay secret.

Verifying a real proof

bin/cp_verify CONFIG_DIR PUBLIC_KEY_DIR REQUEST_FILE TOKEN_FILE PROOF_FILE -id COMMITMENT_ID
Exit Last line Meaning
0 YES / WITHIN LIMITS: … Accepted under exact match / clipped gap
1 NO: <name> Rejected (or a local fault)
5 NOT CHECKED: <name> No verdict; retry once only for internal-stage or internal-exception-parFor, otherwise report
14 NOT-A-VERIFICATION The call was invalid; see the lines after ---- the call ----
2, 3, 4, 15 Bad command line or file, or a partial call; no verdict

verifier/docs/output.md has the full table. integrations/check_proof.py wraps the call and the comparisons.

Tests

D=$(mktemp -d) && ./run_suite.py cpu "$D/work"

These are the CPU tests. They overwrite bin/cp_verify and bin/weight_sums.

See SUITE.md.

Rules


More

Licence and acknowledgments

Apache 2.0: see LICENSE. Material from other projects keeps its own licence. NOTICE and THIRD_PARTY.md list it, PUBLISHED-WORKS.md lists the published works that the code applies or names, and licenses/ has the texts.

We thank Alessandro Chiesa, Giacomo Fenzi, Guy Weissenberg, Miranda Christ, Rahul Dalal, and Ben Mayer for their input and collaboration on FlashProve. We also thank James Petrie for VerInf.

FlashProve (anthropics/FlashProve)

正在搜索数据库…

这是什么?一个研究原型,让 AI 服务提供方能够证明某次请求是由哪个模型回答的,同时不透露模型的权重。
这是我可以调用的服务,还是需要安装的东西?都不是,严格来说。它是一个研究产物:一个可以在本地构建和运行的 CPU 验证器(cpverify),外加一个 GPU 证明器。README 中说明该项目已不再维护。
怎么让它跑起来?以普通用户身份在仓库根目录下运行所有命令,绝不能以 root 身份或通过 sudo 运行。最快的检查方式是构建 cpverify 并验证已存储的示例证明,构建大约需要 1-2 分钟。
开始前需要准备什么?用于验证:Linux、bash、g++ 13 或更高版本,以及内存低于 2 GB。README 还提到,只有在需要安装系统软件包时,才应向用户申请 root 权限。
能直接投入实际使用吗?不能。它是一个已停止维护的研究原型,不接受贡献。README 链接到一份单独的已知限制清单,位于 docs/limits.md。
证明会泄露权重吗?不会。所有依赖权重的已公布数值都会被遮蔽,仅事先声明的列表除外,因此权重保持保密。
试用是否需要 H100?验证已存储的证明不需要,它在普通 CPU 上只需几秒钟即可完成。证明则需要:证明器要在单块 NVIDIA H100 上运行 Llama 4 Maverick。
谁可以检查证明,需要什么?任何人都可以在普通 CPU 上用 cpverify 几秒钟内检查证明,无需权重。
承诺 ID 有什么作用?服务提供方会公布一个承诺 ID,将自己绑定到一组特定的权重,此后的每个证明都必须与这个已承诺的模型相符。
我能看到背后的数学原理吗?可以。lean/THEOREM.md 中有该定理,并说明了如何检查它。
它证明的是哪个模型?完整尺寸的 Llama 4 Maverick,上下文窗口为 1,024 个 token,运行在单块 NVIDIA H100 上。
大小
170.4 MB
星标
1.1K
更新
提交Oct 30
时长
19天
提交 · 90 天