INDEX

Explanations

blank or whitespace characters

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

code and markup snippets

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

 Konsequ

-0.36

 Tourismus

-0.34

principalColumn

-0.34

 başına

-0.34

yczą

-0.33

 Beispiele

-0.33

yarnpkg

-0.33

 Tenden

-0.32

 Werbe

-0.32

 Ausbau

-0.31

POSITIVE LOGITS

TagMode

0.72

<?

0.62



0.61

 lateinit

0.59

RAM

0.58

FormTagHelper

0.57

errHandler

0.56

SAD

0.56

SFD

0.55

뀐

0.55

Activations Density 0.004%

blank or whitespace characters

code and markup snippets

No Comments

No Known Activations