Get Started
Research questionHow can code-generating agents plan programs that satisfy memory and runtime budgets?Without runtime and memory limits, code-generating agents may choose implementations that exhaust resources or miss execution deadlines. These constraints can affect data representation, traversal, and buffering choices before the program runs.
AI Agents
Code Generation & Program Synthesis
Latest papersRecent research connected to this question, newest first.Substrate-Aware AI Agents: Execution Context as a First-Class InputThe evidence concerns high-dimensional pairwise Euclidean-distance programs generated from either the task alone or explicit 128 MB/10-second and 96 MB contracts. It compares three named frontier model configurations and reports structural code changes, memory, wall time, and budgeted correctness; the results are a controlled proof of concept rather than evidence across all agent tasks.research paper · Sep 4, 2026
Related questions
How can agent runtimes avoid context poisoning and latency from growing histories during long-horizon skill execution?How can coding agents reliably implement systems-level requirements and detect the defects they introduce?How can long-horizon LLM agents learn when to group actions without overcommitting?How can AI agents adapt execution routes as runtime evidence invalidates their planned continuation?
Home
Topics
Search
Library