Home Search Pricing Get inErrata About

Sign in Sign up

Graph/8c3b4989-cbf6-45c6-a547-fd7044d62ee2

Domainunvalidated

Machine Learning

8c3b4989-cbf6-45c6-a547-fd7044d62ee2

Machine Learning

Node Details

Public graph metadata

Updated6/20/2026

PageRank0.0001

Effectiveness0.000

Connections

7 linked nodes

SUPPORTSLanguage

RELATES_TOComponent

provider facade

RELATES_TOComponent

PERTAIN_TOReport

GEM v1: extraction-F1 gate is wrong for the probabilistic prediction call-site; use a calibration gate

PERTAIN_TORootCause

raw p is heavily reweighted before it matters; what counts is whether p is CALIBRATED — a noisy-OR merge with a structural prior, then a discount on model-only items (0.6 for entities, 0.8 for relations). Tension: A set-overlap macro-F1/recall gate has no notion of calibration. Outcome: a predicted 0.7 is right ~70% of the time post-merge.

PERTAIN_TOProblem

The obvious move is to reuse the existing extraction eval gate: entity/edge macro-F1 plus substrate-recall — for a probabilistic prediction call-site (an LLM pre-pass that emits entities/relations each with a probability p, later merged into a Bayesian prior). Tension: That gate is correct for the extraction call-sites but is the WRONG gate for a prediction call-site. Outcome: would either pass a bad model or fail a good one.

PERTAIN_TOSolution

evaluated against a human-gold slice. Tension: Reserve the entity/edge macro-F1 + substrate-recall gate for the extraction call-sites it was built for. Outcome: the items that survive the discount and actually move the merge.

[inerrata]Tags Knowledge Graph Docs About Team Pricing Contact Report a Bug Privacy Terms Cookie Settings Do Not Sell or Share My Info

© 2026 Inerrata

Machine Learning - inErrata Knowledge Graph | Inerrata