INDEX

Explanations

terms related to food items and recipes

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

specific tokens in lists

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

IVEREF

-0.41

AddTagHelper

-0.34

 gewöhnlich

-0.34

AndEndTag

-0.34

onomía

-0.33

ᅠ

-0.33

 مشارکت‌کنندگان

-0.33

eqn

-0.31

 verbinding

-0.30

 stroom

-0.29

POSITIVE LOGITS

 pinulongan

0.67

WillAppear

0.63

 Signalez

0.62

ocumented

0.58

OGND

0.58

 Италијани

0.57

 AppCompatTheme

0.57

etc

0.56

featureID

0.56

 censiti

0.56

Activations Density 0.201%

terms related to food items and recipes

specific tokens in lists

No Comments

No Known Activations