liveBook · Manning

Rationale signal

As explained in Retrieval Augmented Generation, The Foundational Ideas

A rationale signal is information passed to a decoder that indicates how strongly passages support an answer.

A rationale signal tells a decoder which retrieved passages contain justification for an answer. Rather than treating every passage equally, it supplies confidence scores that guide attention toward supporting evidence and away from related noise. section 5.6.1

Ten passages, three useful sources

A standard FiD model may receive ten passages when only three contain the answer’s evidence. RFiD trains the model to identify those rationale-bearing passages, then passes their scores to the decoder so generation relies more heavily on them. section 5.6.1

The signal supports two linked goals:

  • Faithfulness: ground the generated answer in passages judged to contain its justification. section 5.6.1
  • Interpretability: expose which passages the model trusted most. section 5.6.1
  • Joint learning: optimize answer generation and rationale classification through a multi-task loss. section 5.6.1

How Retrieval Augmented Generation, The Foundational Ideas covers it

How deep it goes
Explains confidence scoring, decoder guidance, interpretability, and joint training, but does not detail the scoring architecture.
Examples
Uses a ten-passage scenario in which three passages contain evidence and seven contribute related noise.
What you can do after
Apply rationale scores to make generated answers more evidence-grounded and inspect which passages support them.
How it connects
Connects passage retrieval, decoder attention, answer generation, faithfulness, interpretability, and multi-task learning.

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.