An evaluator runs a policy pair under closely matched starting conditions, then supplies preference, progress and explanatory feedback. Conditions may differ across pairs by design.
RoboArena research team · Publication date not disclosed · Source accessed 2026-09-28
Reading results
The paper studies task-aware aggregation of preferences as well as familiar ranking methods. Such a ranking describes its tested policy pool and task distribution; it is not a universal capability score.
robo-arena · Publication date not disclosed · Source accessed 2026-09-28
Official README inspected live. Repository code terms do not automatically cover model weights, data, or third-party assets.
Reported benchmark results
Context before scores. Results from different benchmarks are not directly comparable. A simulation result does not establish real-world reliability, safety or commercial availability.
No result meets our complete revision and methodology requirements for this record yet. Inspect the original benchmark documentation before comparing published scores.
What this evidence does not establish
Human preference and task sampling affect results. Simulation preflight instructions are separate from the real-robot benchmark.
Related reading is an editorial crosslink. Sourced connections describe relationships reported in the cited material. A link to a versioned profile does not establish compatibility with that version unless the connection note explicitly identifies it.