Skip to content
EN

‹ Culture 3.14 · What we ask ourselves

Question · Information

How does a system know it has found enough evidence?

A system can retrieve ten relevant passages and still have nothing to support the claim it is about to write.


Why it matters

Retrieval measures similarity, not sufficiency. A search engine returns whatever is closest to the question, even when the closest thing is not enough to answer it.

When the system writes the answer anyway, the result is correct in form and unfounded in substance. It is the hardest failure to spot by reading the output.

What we know so far

It works reasonably well to require every claim to be anchored to a specific passage, and to discard the claim when the anchor does not appear. It also helps to split the question into sub-questions and check the coverage of each one separately.

Score thresholds, on the other hand, have behaved worse than we expected: they shift with the corpus and with the language.

What remains open

We do not know how to distinguish “there is no evidence” from “the evidence exists but we have not retrieved it”. For the reader, both situations produce the same abstention, and they do not deserve the same answer.

It also remains open how much evidence is enough when the question asks for a synthesis rather than a fact.

Write to the team

Search · Type to search. Esc to close.

Search results →

Ask the bot

This is Ask 3.14, an automatic assistant. It answers only with the public content of this website — solutions, capabilities, systems, questions, readings and laboratory work —, shows the content it has used and distinguishes what the website does not yet allow it to establish. It is not a person from the team and it does not know your case; to speak with someone, write to the team.

Automatic assistant limited to the public content of 3.14; it may abstain. To speak with a person, write to the team.

Write to the team →