INDEX

Explanations

power

np_max-act · gemini-2.0-flash

This neuron selectively activates on numeric or numerical‐measurement tokens (e.g. decimal values, percentages, units) in the text.

oai_token-act-pair · o4-mini Triggered by @xinyanhu8

New Auto-Interp

Configuration

andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_11/trainer_1

Dataset (Dashboard)

Various

Features

131,072

Data Type

float32

Hook Name

blocks.11.hook_resid_post

Architecture

standard

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

одерж

-0.07

 brains

-0.06

YES

-0.06

 morally

-0.06

roy

-0.06

ationally

-0.06

_block

-0.06

xpath

-0.06

acity

-0.06

onClick

-0.06

POSITIVE LOGITS

 Ciudad

0.07

 هناك

0.06

uses

0.06

 Morm

0.06

 forme

0.06

 buena

0.06

 mekan

0.06

енных

0.06

_tF

0.06

�

0.06

Activations Density 0.054%

power

This neuron selectively activates on numeric or numerical‐measurement tokens (e.g. decimal values, percentages, units) in the text.

No Comments

No Known Activations