SGI Deep Research

Scientific general intelligence deep-research subset.

SGI Deep Research is the AgentCompass subset for scientific general intelligence workflows aligned with scientist-style evaluation.

Runtime Status

Field	Value
Benchmark id	sgi_deep_research
Tags	Deep Research, Scientific Research, Judge-Scored
Execution type	local with judge
Typical harness	researchharness
Typical environment	host_process
Current status	registered in the direct runtime

Run Example

agentcompass run \
  sgi_deep_research \
  researchharness \
  your-model \
  --env <env-provider> \
  --benchmark-params '{"sample_ids":["<task-id>"]}' \
  --model-base-url "$MODEL_BASE_URL" \
  --model-api-key "$MODEL_API_KEY"

Common Parameters

Parameter	Meaning
category	Filter task category.
judge_model	Judge model spec for scoring.

Outputs

Per-task details are written to results/sgi_deep_research/<model>/<run>/details/. Aggregate metrics are written to summary.md.