INDEX

Explanations

negations or negative statements

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

wildcard, certificate, course

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

 biais

-0.39

 couverts

-0.36

 UITableViewCell

-0.33

Amerika

-0.32

DeleteBehavior

-0.32

 urbanas

-0.32

 contorno

-0.32

 zimowa

-0.32

 gimnas

-0.32

 helicópter

-0.32

POSITIVE LOGITS



0.61

 المعيارى

0.59

matchCondition

0.58

EDEFAULT

0.58

 للمعارف

0.57

hyrchwyd

0.56

 nahilalakip

0.56

 otomatig

0.55

LookAnd

0.54

➋

0.53

Activations Density 0.023%

negations or negative statements

wildcard, certificate, course

No Comments

No Known Activations