RAG context pruning with sequence labeling
Provence classifier prunes irrelevant context from RAG by formulating as sequence labeling; achieves 80% compression with 2x inference speedup
AA clean typed decision (yes/no, a pick from a list, or a level on a scale) that is concrete, repeats routinely, and has meaningful value.4.24
Key facts
- Vertical
- Software & tech
- Function
- Agents & dev
- Status
- Seen in the wild
- Volume
- routine
- Value
- meaningful
- Risk
- moderate
- Evidence
- built and shown
- Flags
- check-fit
Source: https://arxiv.org/pdf/2501.16214
Build this with a classifier
Define a typed decision with a bounded answer, then evaluate it on examples.
{
"decision_type": "choice",
"question": "Does this input match the decision in “RAG context pruning with sequence labeling”?",
"input": "<input to classify>",
"output": "one label from a fixed list"
}Related use cases
Cite this
Copy a link in your preferred format.