This paper demonstrates that multilingual code-switching instruction tuning, involving four languages (English, Japanese, Korean, and Chinese), significantly improves average multilingual performance,…

This paper proposes a reinforcement learning approach for adapting audio-language models to code-switched speech using group relative policy optimization and verifiable rewards.

The paper quantitatively confirms the Currier A/B language distinction in the Voynich Manuscript, demonstrating it is governed by a higher-dimensional, context-dependent boolean switch rather than a s…

The paper introduces CodeGolf Bench, a novel multi-language benchmark using code golf to measure LLMs' ability to generate highly concise and efficient code, showing that reasoning models significantl…

Minor, single-character perturbations to prompts can significantly degrade the security of code generated by LLMs, suggesting that prompt fragility is a major security concern beyond simple prompt inj…

The paper introduces functional entropy, a code-specific uncertainty quantification method, which successfully predicts functional correctness in LLM-generated code by replacing natural language seman…

The paper introduces Spaghetti Architect, a tool that generates controlled code datasets for machine learning models by deliberately adding redundancy, messiness, and difficulty.

Elevator is a novel, deterministic binary translator that statically translates entire x86-64 executables to AArch64 by considering all possible interpretations of every byte, eliminating the need for…

The paper introduces Poison-with-Style (PwS), a stealthy model poisoning attack that exploits developers' inherent code styles as covert triggers to make Code LLMs generate vulnerable code without exp…

The paper proposes Fresco, a novel bidirectional type inference approach for first-class polymorphism that enables local type information to flow back and forth between functions and arguments.

This paper proposes source code algebra as an alternative to text editing for modifying source code, allowing agents to make semantic changes with fewer tokens and higher success rates.

FPMoE introduces a sparse Mixture-of-Experts (MoE) architecture to improve functional code generation across multiple functional programming languages, achieving state-of-the-art performance with fewe…

This paper analyzes various source-to-bytecode obfuscation techniques for Erlang, demonstrating that effective protection relies on exploiting the representational gaps between high-level semantics an…

This paper introduces TRIM, an algorithm that minimizes redundant edits in AI-generated code, called CodeSlop, by minimizing agent trajectories, reducing CodeSlop by 17.9%-32.9% with negligible perfor…

This study analyzes over 20,000 real-world coding sessions to show that AI coding agents frequently fail users through subtle misalignment, requiring constant manual correction even when major system…

MATRIX is a novel, robust code watermarking framework that encodes watermarks using constrained parity-check matrix equations, achieving high detection accuracy and improved robustness for code proven…