> Tiered Verification and Red-Team Attacks on LLM Inference

Zero-Knowledge Proofs 高级 2026-04-01 22:52 2026-04-01
#verification #security #red-team #weight poisoning #model substitution #ZK proofs #tiered security

A three-tier verification framework (SHA-256 commits → Fisher-selected proofs → full ZK) with 1000x cost difference between tiers. Red-team finding: 0.1% weight poisoning produces KL=0.00 — identical output — but injects backdoors. ZK is the only defense.

Tiered Verification and Red-Team Attacks

Three-Level Verification Framework

Tier Method Cost Guarantee Use Case
1 SHA-256 commitment chain ~$0.001 Trust-based 99% of queries
2 Fisher-selected layer proofs ~$0.10 Probabilistic Disputes
3 Full ZK proof ~$2.23 Cryptographic Blockchain AI markets

1000x cost difference between tiers. Optimistic model: assume honesty, escalate on dispute with slashing.

Red-Team Attack Detection

Tested on LLaMA-3.2-1B (1.24B params):

Attack Output Normal? Analytics Detect? ZK Detects? Mechanism
Model substitution (160M for 1B) Yes No Yes Weight commitment
Layer skipping (8/16 layers) Partial No Yes Commitment chain
Weight poisoning (0.1%) Yes (KL=0.00!) No Yes Merkle root

The Weight Poisoning Threat

This is the most dangerous attack: modifying 0.1% of weights produces identical output (KL divergence = 0.00) on standard inputs, but can inject arbitrary backdoors activated by specific trigger patterns.

  • Output-based monitoring: useless (zero statistical divergence)
  • Perplexity monitoring: useless (unchanged)
  • ZK weight commitment: the only cryptographic defense

The commitment chain catches this because the Merkle root of model weights changes, even if only 0.1% of parameters are modified.

Strategic Insight

From the HanFei system design:

"Users only care about one thing: Did you run the correct model?"

Three realizations: 1. Level 2 (signature + re-run) covers 90% of real deployment needs 2. Level 3 (full ZK) is only necessary in decentralized AI markets (blockchain) 3. Providing GPU MSM infrastructure (hanfei-shu) is strategic positioning for when this market matures

The Long-Sequence Challenge

For LLaMA-3 at 8192 tokens, attention requires 275B Fp multiplications per layer — infeasible to prove directly.

Solution: Chunked sequence proving (256-token chunks) reduces attention ops by 1024x while maintaining compositional soundness through inter-chunk commitment linking.

Scaling to Production

The HanFei ecosystem roadmap: - hanfei-shu (术, technique): GPU MSM crate — published on crates.io - hanfei-shi (势, momentum): IPA proving backend — packaging Q2 2026 - hanfei-fa (法, method): Full ZK framework — Q3 2026

Each crate is independently useful, creating adoption momentum bottom-up.