INDEX

Explanations

phrases requesting additional information or data

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

click here for more

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

IUrlHelper

-0.71

/*

-0.68

awtextra

-0.65

ⓧ

-0.64

 Италијани

-0.63

 itſelf

-0.62

 ſta

-0.60

 ſtre

-0.60

webElementXpaths

-0.60

 pleaſure

-0.58

POSITIVE LOGITS

the

0.45

0.43

 actual

0.41

 official

0.41

 coffee

0.40

 electronic

0.39

けて

0.38

 Electronic

0.38

 edit

0.38

 page

0.38

Activations Density 0.011%

phrases requesting additional information or data

click here for more

No Comments

No Known Activations