Strange Loop data
Researcher Trajectories
A growing collection of authentic human–agent research records. Each trajectory preserves the question, agent work, researcher interventions, evidence, and outcomes that are usually lost between a project's first experiment and its final paper.
What this data captures
Research is a sequence of decisions under uncertainty. Rather than recording only the polished outcome, these records retain the moments that shape the work: a proposed plan, a redirected experiment, a failed run, or a judgment about what to do next.
Source trajectory
The research goal, context, tools, actions, and artifacts from a real human–agent collaboration.
Complete data record
Every trajectory is presented as one complete research record, rather than split into separate data categories.
- Trajectories
- 8
- Research areas
- 4
Browse trajectories
Search the current collection of research records.
- Long-Horizon Offline GCRL Value Errorsgcrl~46.7K tokens
- Can a decoding method that measures novelty in a …Title truncated in the source doc — replace with the full research question.llm~42.2K tokens
- Is there a characteristic geometric structure tha…Title truncated in the source doc — replace with the full research question.rl~40.2K tokens
- How close to optimality are current state-of-the-…Title truncated in the source doc — replace with the full research question.rl~38.1K tokens
- Augmentor Policy Handoffsrl~36.5K tokens
- RL For Novel LLM Capabilitiesllm~32.5K tokens
- CRL Importance Samplingcrl~22.5K tokens
- Analogical RL Representation Comparisonrl~20.1K tokens