AI Paper Atlas

Research map / Large language models / Agents, code and math

Not All Invariants Are Equal: Curating Training Data to Accelerate Program Verification with SLMs

Ido Pinto, Yizhak Yisrael Elboher, Haoze Wu, Nina Narodytska, Guy Katz

ICML 2026 · Large language models › Agents, code and math · approach: Execution-based code verification · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add LLM-driven simplification and augmentation (V2) on top of AST normalization (ablation)Qwen3-4B · InvBench Hard (n=123)R_correct (%): 33.1 -> 44.4Improved
add LLM-driven simplification and augmentation (V2) on top of AST normalization (ablation)Qwen3-8B · InvBench Hard (n=123)R_correct (%): 30.1 -> 42.8Improved
add LLM-driven simplification and augmentation (V2) on top of AST normalization (ablation)Llama-3.1-8B · InvBench Hard (n=123)R_correct (%): 36.3 -> 45.5Improved
add LLM-driven simplification and augmentation (V2) on top of AST normalization (ablation)Mistral-7B · InvBench Hard (n=123)R_correct (%): 24.9 -> 31.4Improved
add AST normalization (V1) to raw verifier invariants (ablation)Qwen3-4B · InvBench Hard (n=123)R_valid (%): 81.3 -> 97.6Improved
fine-tune on raw UAutomizer invariants (V0) instead of using the base model (ablation)Qwen3-4B · InvBench Hard (n=123)R_valid (%): 99.2 -> 81.3Hurt
fine-tune on raw UAutomizer invariants (V0) instead of using the base model (ablation)Mistral-7B · InvBench Hard (n=123)R_valid (%): 93.8% -> 65.0%Hurt
fine-tune on raw UAutomizer invariants (V0) instead of using the base model (ablation)Qwen3-0.6B · InvBench Hard (n=123)R_correct (%): 28.5 -> 18.7Hurt
fine-tune on raw UAutomizer invariants (V0) instead of using the base model (ablation)Llama-3.1-8B · InvBench Hard (n=123)R_correct (%): 14.9 -> 31.2Improved
fine-tune on Wonda-curated data (V2) instead of the base modelQwen3-4B · InvBench Hard (n=123)R_speedup (%): 11.1 -> 24.7Improved
fine-tune on Wonda-curated data (V2) instead of the base modelQwen3-0.6B · InvBench Hard (n=123)R_correct (%): 28.5 -> 27.9No clear effect
fine-tune on Wonda-curated data (V2) instead of the base modelQwen3-0.6B · InvBench Hard (n=123)VBP_E2E (s): 183.0 -> 174.1Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy