INDEX

Explanations

references to numerical data or statistics related to groups

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

numbers and times

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

Sig

-0.37

rook

-0.35

Bronnen

-0.34

ensburg

-0.34

 inden

-0.33

 asynchronous

-0.32

vir

-0.32

UNICODE

-0.32

 trom

-0.32

 Văn

-0.32

POSITIVE LOGITS

 nahilalakip

0.68

 vždy

0.67

 jeweils

0.66

 zawsze

0.64

Бахар

0.63

MLLoader

0.61

RenderAtEndOf

0.61

 Siempre

0.57

 תמיד

0.56

 siempre

0.55

Activations Density 0.302%

references to numerical data or statistics related to groups

numbers and times

No Comments

No Known Activations