> ## Documentation Index
> Fetch the complete documentation index at: https://docs.kalarislabs.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Machine learning research skills

> Select agent skills for ML experiments, model evaluation, reproducibility and conference paper writing.

# Machine learning research

Use a skill that matches the stage of your experiment. Keep datasets, training code and metrics as the source of truth; ask the agent to identify missing evidence before it drafts claims.

| Task | Skill | What to provide |
| - | - | - |
| Plan or report a model evaluation | [`evaluating-llms-harness`](/research-agent-skills/skills/evaluating-llms-harness) | Task definitions, evaluation set and scoring rules |
| Train across multiple GPUs | [`pytorch-fsdp2`](/research-agent-skills/skills/pytorch-fsdp2) | Model, hardware, checkpoint and memory constraints |
| Write a conference paper | [`ml-paper-writing`](/research-agent-skills/skills/ml-paper-writing) | Results, baselines, figures and target venue |
| Document repeatability | [`reproducibility-statement`](/research-agent-skills/skills/reproducibility-statement) | Seeds, splits, compute budget, code and data access |

Install one skill into your current project:

```bash theme={null}
npx skills add KalarisLabs/research-agent-skills --skill ml-paper-writing
```

Try: “Use the results in `runs/` and the comparison table to outline an ML paper. Mark every claim that needs another experiment.” Check the current venue requirements before submission. For interpretability work, also see [AI research](/research-agent-skills/fields/artificial-intelligence).


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.