Researchers Introduce MPC-Patch-Bench, First Security-Focused Benchmark for LLM Code Repair in Secure Multi-Party Computation
Researchers have introduced MPC-Patch-Bench, a repository-level benchmark for evaluating how well large language models can repair Secure Multi-Party Computation (MPC) software while preserving cryptographic safety. The benchmark addresses gaps in existing tools like SWE-bench, which fail on MPC codebases due to structural mismatches and an inability to verify cryptographic correctness. The findings show that even the best-performing LLM resolves only 17.1% of tasks when security checks are applied, with up to 40% of functionally passing patches rejected for cryptographic or numerical violations.
A preprint posted to arXiv introduces MPC-Patch-Bench, the first repository-level benchmark designed to evaluate LLM-based code repair specifically for Secure Multi-Party Computation software. The authors argue that general-purpose benchmarks like SWE-bench are structurally unsuited to MPC repositories, which are dominated by generic Python infrastructure rather than cryptographic logic and lack the standardized tests those pipelines require. MPC-Patch-Bench consists of 205 fully verified instances curated through a domain-specific agent that filters pull requests across three cryptographic layers, supplemented by a human-AI engine that synthesizes missing problem statements and test cases. A dedicated MPC Verifier performs dynamic differential testing against plaintext oracles and applies static analysis rules to flag unsafe reveals, insecure arithmetic, and illegal public/private type casts. The strongest LLM evaluated achieved a functional resolution rate of 22.9%, which dropped to 17.1% after the MPC Verifier's security and numerical-fidelity checks were applied. These results highlight a significant gap between surface-level code correctness and the deeper cryptographic guarantees required for MPC deployment in privacy-preserving machine learning, biomedical collaboration, and secure analytics.
What's missing
As a preprint, this work has not yet undergone peer review. It is also unclear how the 205 benchmark instances were distributed across MPC frameworks or how representative they are of real-world MPC codebases. The benchmark currently covers only two frameworks, and the authors do not discuss how performance might scale with more capable future models or fine-tuned domain-specific LLMs.
What different sources said
- arXiv cs.AICenter
MPC-Patch-Bench: Security-Aware LLM Code Patch for Multi-Party Computation
Related
Gut Bacteria Enzyme Found to Break Down Heat-Processed Food Compounds, Producing Novel Biogenic Amines
Researchers have discovered that an enzyme in common gut bacteria can degrade N-epsilon-carboxymethyllysine (CML), a compound formed during thermal food processing, producing previously unknown biogenic amines. The enzyme, ornithine decarboxylase SpeC from enterobacteria, acts on CML and related modified lysine derivatives through a low-level 'underground' catalytic activity. This finding suggests a previously unrecognized communication axis between thermally processed dietary compounds and gut microbial physiology, with potential implications for host health.
Full-Length Gene Sequencing Reveals Two Distinct Bacterial Communities in Black-Legged Ticks Expanding Into Canada
Researchers used Oxford Nanopore full-length 16S rRNA gene sequencing to characterize the microbiome of Ixodes scapularis black-legged ticks collected in Nova Scotia, Canada, distinguishing between tick-adapted bacteria and environmentally acquired bacteria. The study comes as I. scapularis — the primary vector of Lyme disease — is rapidly expanding northward into Canada due to climate change. The findings suggest that environmentally derived bacteria in tick microbiomes are not mere contamination, which has implications for how tick microbiome data is collected and interpreted across surveillance studies.
Study Identifies Metabolic Link Between Cell Envelope Stress and Biofilm Formation in Bacteria
Researchers have discovered that the metabolite acetyl-CoA directly inhibits enzymes that degrade the bacterial signaling molecule c-di-GMP, connecting cell envelope biosynthesis stress to biofilm formation in Pseudomonas aeruginosa. The study found that sub-inhibitory concentrations of antibiotics targeting early peptidoglycan biosynthesis — but not other antibiotic classes — elevate c-di-GMP levels by reducing phosphodiesterase activity, with acetyl-CoA competing for the enzyme active site. Because the relevant enzyme domain is broadly conserved across bacterial species, this checkpoint mechanism may be widespread and could have implications for understanding antibiotic-induced biofilm responses.