Ikibazo gihuza ibibazo · Hagati Urwego
Ikibazo cya Scaffolding hamwe na Harnesses Ikibazo
Gerageza umukozi woherejwe hamwe no kohereza ibikoresho, guhagarika amategeko no gukemura amakosa, n'impamvu icyitegererezo kimwe gitanga amanota atandukanye murwego rwo hejuru.
Ikibazo 1 Bya 8
Niki ibikoresho bya agent?
Reba buri kibazo8 ibibazo
Ibisubizo nibisobanuro bigaragara mugihe ufata ikibazo.
- Niki ibikoresho bya agent?
- The model's neural network weights
- The code around a model that runs the loop, executes tools and decides when to stop
- A dataset for training agents
- Ubwoko bwa GPU
- Ni uruhe rugero rw'amategeko ahagarika?
- Using a larger model
- A maximum number of steps or a cost budget
- Adding more tools
- Changing the font of outputs
- Umushinga SWE-agent washimangiye iki?
- That models need no tools
- An agent-computer interface designed for models, with concise viewers and immediate feedback
- That bigger GPUs solve all agent problems
- Removing the loop entirely
- Kuki icyitegererezo kimwe gishobora gutsinda amanota atandukanye kuri SWE-intebe muri raporo zitandukanye?
- The benchmark changes every hour
- Scores reflect the model plus harness, tools, step limits and retry policy
- Models forget their training between runs
- Scores are randomly assigned
- Nigute ibikoresho byakagombye gukemura ikibazo cyatsinzwe?
- Crash the entire run immediately
- Return the error to the model as a tool result it can react to
- Silently ignore it
- Restart the computer
- What is the purpose of capping tool output size and marking truncation?
- To hide information from users
- To keep context manageable and let the model know to search or page for more
- To make tools run faster on the GPU
- To prevent the model from using tools
- When fairly comparing two models as agents, what should you hold constant?
- Only the model's name
- The harness, prompt, tools, step limit and retry policy
- Nothing; vary everything
- The date of the test only
- A harness pauses before large refund tool calls for human approval. What design principle does this show?
- Context compaction
- Gating side-effecting tools with permissions or confirmation
- Parallel function calling
- Embedding fine-tuning