INDEX

Explanations

This neuron identifies topics related to American history, politics, or culture as well as military terms and places.

oai_token-act-pair · gemini-2.0-flash

academic subjects or disciplines

np_token-act-pair-logits · gpt-4o-mini

Fields of study

np_max-act-logits · gemini-2.0-flash

say "topics" Method used: 2 — tokens after indicate topical/subject words — say "topics"

np_max-act-logits · gpt-5-mini Triggered by @chenshw0109

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_21/width_16k/average_l0_13

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.ln2.hook_normalized

Architecture

jumprelu_transcoder

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 fallait

-0.64

 épreuve

-0.50

 része

-0.49

httphttps

-0.49

 cumplido

-0.48

 fidélité

-0.47

 wacht

-0.47

UnusedPrivate

-0.46

 fût

-0.45

 planète

-0.45

POSITIVE LOGITS

 medicine

1.17

 economics

1.16

 theology

1.14

 biology

1.13

 politics

1.09

 astronomy

1.09

 pharmacology

1.09

 geology

1.08

 dentistry

1.07

 photography

1.05

Activations Density 5.297%

This neuron identifies topics related to American history, politics, or culture as well as military terms and places.

academic subjects or disciplines

Fields of study

say "topics" Method used: 2 — tokens after indicate topical/subject words — say "topics"

No Comments

No Known Activations

This neuron identifies topics related to American history, politics, or culture as well as military terms and places.

academic subjects or disciplines

Fields of study

say "topics" Method used: 2 — tokens after indicate topical/subject words — say "topics"

No Comments

No Known Activations