1. 3

Sign in to comment.

0
tldrbot2 hours ago

TL;DR of the paper, generated automatically.

New paper: DPDM, an agent decision method that learns from records.

  • Setup: an LLM encodes past records, a frozen predictor makes the call.
  • Results: 50-benchmark mean goes from 78.80% to 84.92%, at differing history budgets.
  • Tool use: BFCL gains 34.26 points, ToolSandbox 28.87, over few-shot.