INDEX

Explanations

race

np_max-act · gemini-2.0-flash

The neuron primarily detects mentions of “race” (and its variants) in event names and contexts, especially when paired with race numbers.

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

 Piet

-0.07

 Doming

-0.07

 Bias

-0.06

 Wells

-0.06

Gol

-0.06

.setLayoutParams

-0.06

融

-0.06

 روان

-0.06

:X

-0.06

 airy

-0.06

POSITIVE LOGITS

rv

0.07

าคาร

0.06

xce

0.06

يكي

0.06

 Sabbath

0.06

()};↵

0.06

inton

0.06

 features

0.06

mqtt

0.06

[list

0.06

Activations Density 0.007%

race

The neuron primarily detects mentions of “race” (and its variants) in event names and contexts, especially when paired with race numbers.

No Comments

No Known Activations