PaLM scales to 540B parameters with reasoning gains

Monday 4 April 2022 · significance ●●○○○

Google's PaLM shows breakthrough performance on multi-step reasoning with chain-of-thought prompting.

What happened and why it matters

PaLM, trained on 6,144 TPU v4 chips, combined scale with chain-of-thought prompting to produce marked jumps on arithmetic and commonsense reasoning, and could explain jokes — capabilities absent in smaller models.

Sources

Pathways Language Model (PaLM): Scaling to 540 Billion Parameters for Breakthrough Performance Google Research · 4 Apr 2022

Effect on the Gapnium Coefficient

+0.0030

0.1118 → 0.1148

Adjustments

gapnium.cognition +0.012 Multi-step reasoning emerges with scale plus chain-of-thought.
gapnium.compute +0.006 Pathways system demonstrates efficient training across thousands of accelerators.

Review

Last reviewed by AI agent: 7 Nov 2026
Model: seed:claude-opus-5-5
Agent confidence: 0.75
Human reviewed: —

Initial historical seed. Scores assigned by the methodology rubric; awaiting first scheduled re-review.

Feedback on this scoring

Hover the chart to read values · click a marker to open its articles