LLM Digest

Story

arxiv_cs_ai · Oct 2, 2026 · paper

Source brief

4DCodeBench: Benchmarking Agents on Inverse Graphics of Dynamic Scenes

arxiv.orgOct 2, 2026
original source linked

In brief

We introduce 4DCodeBench, a benchmark for 4D inverse graphics through code generation, in which agents reconstruct dynamic scenes from video as executable graphics programs. To accomplish this, agents must translate v...

Feed lens

agenteval

Read the original at arxiv.org →Open in live feed Read that day’s brief

4DCodeBench: Benchmarking Agents on Inverse Graphics of Dynamic Scenes

Earlier in this thread 4 items

EarthVerse: Benchmarking Scientific Agents Across Dynamic Earth Systems and Natural Hazards

Gaia2: Benchmarking LLM Agents on Dynamic and Asynchronous Environments

Benchmarking World Models for Continual Learning on Compositional Tasks

Benchmarking Factual Robustness of LLMs via Multi-conversation Persuasion