INDEX

Explanations

references to the English language and associated topics

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 caufe

-1.05

 raiſ

-1.05

 fhew

-1.03

 cauſe

-0.97

=$?

-0.94

المناصب

-0.92

دانشنامهٔ

-0.92

 consultato

-0.91

 lenker

-0.91

 purpoſe

-0.89

POSITIVE LOGITS

 English

2.61

English

2.23

 ENGLISH

2.07

 english

2.01

ENGLISH

1.82

english

1.77

 engl

1.34

 Engl

1.33

 anglais

1.29

 inglés

1.26

Activations Density 0.090%

references to the English language and associated topics

No Comments

No Known Activations