Question Answering on LoCoMo 200-question
19.2F1AURA Intent
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| AURA IntentCondition=full AURA pipeline with IntentInferrer, Backbone=gpt-4o-mini, Seed=12026.06 | 19.2 | 2.5 | 6.19 | 1.42 | 0 | |
| NoIntentCondition=ReAct-style tool use, Backbone=gpt-4o-mini, Seed=12026.06 | 17.3 | 2.5 | 2.91 | 1.33 | 0 | |
| LiteralCondition=single LLM call with session catalog only, Backbone=gpt-4o-mini, Seed=12026.06 | 4.2 | 2 | 1.16 | 0 | — |