Research
WorkBench agents close the workplace reliability gap
WorkBench agents now solve 89 percent of workplace tasks with 2.5 percent harmful actions, changing the risk math for builders.
2 stories tagged open-models.
WorkBench agents now solve 89 percent of workplace tasks with 2.5 percent harmful actions, changing the risk math for builders.
Tiny checkpoint gains become harder to mistake for progress with olmo-eval, Ai2's open workbench for iterative LLM evaluation.