INDEX

Explanations

mathematical symbols and expressions related to matrix operations

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

mu followed by closing punctuation

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

OGND

-1.13

IntoConstraints

-0.84

bootstrapcdn

-0.80

 Exactos

-0.77

TagMode

-0.76

Personensuche

-0.75

 وتسجيلات

-0.74

帖最后由

-0.71

elfare

-0.70

ロウィン

-0.70

POSITIVE LOGITS

0.41

 anyway

0.39

↵↵

0.37

 again

0.34

 nevertheless

0.34

setzung

0.33

 anyways

0.33

 douce

0.32

 nonetheless

0.32

The

0.31

Activations Density 0.009%

mathematical symbols and expressions related to matrix operations

mu followed by closing punctuation

No Comments

No Known Activations