INDEX

Explanations

HTML tags and elements associated with user interface components

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

containing structured data

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

 للاسماء

-0.87

 otomatig

-0.82

IntoConstraints

-0.77

 tartalomajánló

-0.74

 queſta

-0.73

GEBURTS

-0.71

Билгалдахарш

-0.69

 ſche

-0.69

лтамалар

-0.68

featureID

-0.68

POSITIVE LOGITS

 contains

0.39

 containing

0.39

 which

0.36

 consisting

0.36

 area

0.36

 contain

0.34

 chứa

0.34

 Containing

0.33

0.32

 inside

0.32

Activations Density 0.007%

HTML tags and elements associated with user interface components

containing structured data

No Comments

No Known Activations