liveBook · Manning

Self-critique

As explained in Retrieval Augmented Generation, The Foundational Ideas

Self-critique is a model technique in which generated content or intermediate decisions are evaluated by the model itself.

Self-critique lets a model evaluate retrieved evidence and its own generated answer instead of treating retrieval and generation as passive, fixed steps. In Self-RAG, the model can retrieve passages when needed, skip retrieval when its parameters already contain the answer, and reflect during generation. section 9.0

Reflection tokens make the evaluations explicit and controllable at inference time. IsRel filters irrelevant passages, IsSup checks claims against evidence, and IsUse judges whether the resulting answer is worth returning. section 9.8

Self-critique supports two deployment patterns:

  • Prompted Self-RAG emulates reflection tokens with separate grader calls, adding an API round-trip for each critique. section 9.8
  • Trained Self-RAG fine-tunes a smaller model to emit reflection tokens inline, shifting the cost to training rather than repeated critique calls. section 9.8
  • Decoding weights adjust strictness without retraining: higher pressure against unverified statements suits compliance work, while looser settings support creative work. section 9.8

Adaptive answer checking

A query whose answer is already stored in the model can skip search. A query requiring external evidence can retrieve passages, discard irrelevant ones, check claims against them, and reject an answer judged unworthy. The same mechanism can penalize unsupported statements more strongly for compliance than for creative generation. section 9.0 section 9.8

Self-critique is not independent verification: a model judging its own output may share the assumptions that produced it. External evaluators remain necessary when that blind spot matters. section 9.8

How Retrieval Augmented Generation, The Foundational Ideas covers it

How deep it goes
Explains reflection-token roles, prompted and trained implementations, decoding control, and the limitation of shared model assumptions.
Examples
Demonstrates retrieval skipping, irrelevant-passage filtering, claim checking, answer-worthiness judgment, and compliance-versus-creative decoding.
What it assumes
Assumes retrieval-augmented generation with model parameters and external vector-database memory, while explaining reflection-token behavior.
What you can do after
Operate/configure self-critique by choosing prompted or trained reflection and adjusting inference-time strictness.
How it connects
Connects self-critique to retrieval-augmented generation, adaptive retrieval, reflection tokens, and external evaluation.

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Retrieval Augmented Generation, The Foundational Ideas. The section links point to the parts of the book it draws on.