INDEX

Explanations

determiners and articles in the text

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

means or method

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

Còn

-0.57

 itſelf

-0.52

 mijne

-0.51

 sibi

-0.50

 kysy

-0.48

 zijne

-0.45

 nôtre

-0.43

 Theſe

-0.43

 fédé

-0.41

 russes

-0.40

POSITIVE LOGITS

 through

1.05

Through

1.02

 Through

0.98

through

0.97

 THROUGH

0.94

Thru

0.91

 thru

0.90

 Thru

0.85

via

0.83

thru

0.82

Activations Density 0.018%

determiners and articles in the text

means or method

No Comments

No Known Activations