INDEX

Explanations

food-related terms, particularly those describing various cooking methods and types of dishes

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

cooking methods

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

 miniaturka

-0.63

featureID

-0.58

 zoude

-0.54

casila

-0.54

 răsp

-0.54

 dieß

-0.54

asilkan

-0.52

 enfans

-0.51

 przeznac

-0.51

rvore

-0.51

POSITIVE LOGITS

 Baked

0.47

 baked

0.45

 roasted

0.44

bcryptjs

0.41

MP

0.40

baked

0.39

Baked

0.38

 grilled

0.37

 fried

0.36

dried

0.36

Activations Density 0.007%

food-related terms, particularly those describing various cooking methods and types of dishes

cooking methods

No Comments

No Known Activations