Axiomatic Causal Interventions for Reverse Engineering Relevance Computation in Neural Retrieval Models.
Catherine ChenJack MerulloCarsten EickhoffPublished in: SIGIR (2024)
Keyphrases
- reverse engineering
- retrieval model
- test collection
- information retrieval
- relevance feedback
- language model
- language modeling
- statistical language models
- software engineering
- inference network model
- document retrieval
- information retrieval systems
- query topic
- trec collections
- retrieval framework
- query expansion
- retrieval effectiveness
- vector space model
- dynamic analysis
- vector space
- retrieval systems
- gene regulatory networks
- retrieval functions
- language modeling framework
- document collections
- probabilistic retrieval models
- bayesian networks
- reverse engineer
- relevance judgments
- object oriented
- relevant documents
- user queries
- tf idf
- program understanding
- sentence retrieval
- average precision
- active learning
- search engine
- document ranking
- relevance model
- retrieved documents