Code & software

Code generation, coding agents, software engineering with AI, program synthesis

23 articles

All Agents Language models Reinforcement learning Vision Generative media Speech & audio Robotics & embodied AI Safety & security Evaluation & benchmarks Efficiency & systems Code & software AI for science & health Theory

Viewgrid list compact

Training models on runtime program-state reasoning boosts automated software engineering

Visual feedback boosts software engineering agent success rates

Training recipe boosts security in AI-generated code by improving planning and testing

New benchmark tests AI's ability to spot invalid code reviews

Code-linked evidence lets AI software claims expire when dependencies change

Agents learn software repair rules that transfer across repositories

Merged coding-agent performance fixes often miss their claimed gains

Training LLMs to write code boosts medical calculation accuracy by 20-30 points

Query plans make text-to-SQL portable across database dialects

Adoption-aware memory learning boosts LLM agent NPU kernel synthesis

Semantic constraints can prune invalid code during language model decoding

LLM-discovered search operators adapt destruction and repair to search state

One demonstration can bootstrap reusable robot manipulation programs

Reusable skills improve smart-contract audits, but model choice dominates

New benchmark tests LLMs on repository-level reasoning about code execution

AI coding agents still miss production inference failures

Coding agents retrieve better by targeting missing evidence sets

Program repair improves when memories match repository needs and repair stages

Embedded software agents need interaction contracts and continuous assurance

Literate programming links prose, code and symbols for coding agents

Coding agents extend LLM serving simulators with greater fidelity and speed

Switching Models Mid-Task Imposes a Cost and Quality Penalty

Post-training can replace prompt-time schema injection for enterprise coding agents