The paper introduces Tree-like Self-Play (TSP), a novel framework that treats secure code generation as a fine-grained decision process, significantly improving LLM security by forcing the model to se…

The paper introduces HORIZON, a self-evolving agent framework for hardware design using git operations.

This paper introduces Skill Self-Play (Skill-SP), a co-evolutionary framework for LLM training that bridges the gap between structured verification and open-ended exploration.

The paper introduces Grid Programs, a novel, Turing-complete model of computation where programs are two-dimensional arrangements of instructions, fundamentally departing from linear code structures.

The paper proposes Fuzzy-Function Programming and introduces Program-as-Weights (PAW), a compact, locally-executable neural artifact for everyday programming tasks.

This paper surveys 1,250 arXiv papers on self-improving AI systems, categorizing them based on what they improve and the degree of loop closure. It identifies a distinctive feature of self-evaluation…

SEMBridge is a tagless-final framework that allows a single executable object program to generate multiple program semantics, including weakest-precondition and bounded-checking interpretations, ensur…

The paper proposes SRTJ, a Self-Evolving Rule-Driven Training-Free Jailbreak framework that systematically discovers and refines attack strategies using rule composition and feedback to achieve robust…

This paper introduces knowledge-centric self-improvement for AI systems, where agents remain generic and disposable while a curated knowledge base is used for future tasks, leading to more inspectable…

This paper measures the lower bound for the shortest program generating a sequence, proving a conservation law and providing a deterministic engine to recover generating programs for certain sequences…

ASPIRE is a continual learning system that autonomously writes and refines robot control programs in a code-as-policy paradigm, discovering transferable skills and surpassing prior methods on various…

The paper proposes a tamper-proofing model for self-modifying code (SMC) by leveraging external timing, concurrency, and microarchitectural state to make non-SMC reproduction detectably expensive.

This paper introduces InvWeaver, a neuro-symbolic framework for synthesizing loop invariants in programs with multiple interacting loops.

SCOPE introduces a data-free self-play framework that co-evolves a task-generating Challenger and a document-answering Solver, significantly improving open-ended performance on language models without…

This paper proposes using offline reinforcement learning (RL) as an efficient alternative to online RL for post-training code-generating LLMs, demonstrating its effectiveness, especially for smaller m…

The paper presents a tool-making pipeline for production LLM agents that compiles repeated steps into validated, versioned tools before deployment, reducing latency and error rate.

The paper introduces a data-centric optimization pipeline to improve coding agents' ability to interact with a branching lakehouse, showing significant accuracy gains by treating agent evaluation as a…

This paper proposes SkillOpt-Lite, a minimal viable pipeline for skill optimization in autonomous agents, which accelerates convergence and outperforms full SkillOpt.

The OISD framework improves language model reasoning by distilling on-policy predictive signals from the final output layer to intermediate representations, leading to substantial improvements on math…

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…