Perplexity Decider v1.2 decision model explained in technical deep dive
Start here
A technical deep dive published by the author of jbarrow.ai explains how Perplexity Decider v1.2 works: an open decision model built on Qwen-3.8 27B that turns a decoder-only LLM into a non-causal encoder and answers all classification questions in a single prefill step through a 255-option readout head. The write-up describes the /v1/decisions API's yes_no, choice and score question types, training on harmonized datasets such as tasksource with synthetic data and choice shuffling, and post-hoc calibration at temperature T=1.32. The author argues the main case for decision models over autoregressive LLMs is token efficiency, and notes that the 9B Decider Flash variant is trained with a distillation loss.
Generated from reports · updated 1 h ago
Report timeline
Follow the reports to see the event from different sides.
- Hacker News · AIA Deep Dive into Decision Models: How Perplexity Decider v1.2 Works
This post dissects Perplexity Decider v1.2, an open decision model built on Qwen-3.8 27B that converts a decoder-only LLM into a non-causal encoder and answers all classification questions in a single prefill step via a 255-option readout head. It covers the /v1/decisions API's yes_no, choice and score question types, training on harmonized datasets such as tasksource with synthetic data and choice shuffling, and post-hoc calibration at temperature T=1.32. The author argues the main case for decision models over autoregressive LLMs is token efficiency, and notes the 9B Decider Flash variant is trained with a distillation loss.
Heat for this event
Not enough continuous observations yet to draw a trend.