Útmutatóhoz kapcsolódó kvíz · Kemény Szint
Az AI-ügynökök értékelése
Azt teszteli, hogy az ügynök hogyan értékeli a feladat sikerét, a pályákat és az eszközhasználatot, miért számítanak a környezetalapú osztályozásnak és az ismételt próbáknak, valamint az osztályozókkal kapcsolatos kompromisszumokat.
kérdés 1 -ból 8
Why is checking a single final answer usually not enough for agents?
Tekintse meg az összes kérdést8 kérdéseket
A válaszok és magyarázatok megjelennek a kvíz kitöltésekor.
- Why is checking a single final answer usually not enough for agents?
- Agents never produce answers
- Agents act over many steps and cause side effects, so the path and final state matter
- Answers are always correct
- Reference answers cannot be written for agent tasks
- How does SWE-bench judge whether an agent solved an issue?
- By running the repository's tests after the agent's patch
- By counting lines changed
- By asking the agent if it succeeded
- By measuring response speed
- What does a pass^k style metric measure?
- Speed across k tools
- Success on at least one of k attempts
- Whether the agent succeeds on every one of k repeated attempts
- The number of tokens per step
- Which dimension would flag an agent that succeeds but takes fifty steps for a five-step job?
- Task success
- Biztonság
- Tool naming
- Trajectory quality
- What is a known weakness of model-based graders?
- Nem tudnak szöveget olvasni
- They can be biased, for example toward long or confident answers, so they need checking against human labels
- They are always slower than humans
- Csak képeken dolgoznak
- Why reset the environment to a known state before every run?
- So results are reproducible and runs cannot affect each other
- To save disk space
- Because agents require it to start
- To make tasks harder
- Why can public leaderboard scores mislead a team?
- They are always faked
- Leaderboards never publish numbers
- Benchmarks can leak into training data, saturate or fail to match your workload
- They only measure cost
- How should a safety eval treat a run where the agent finished the task but obeyed a planted instruction to leak data?
- As a full success
- As a partial success worth half credit
- It should be ignored
- As a failure