INDEX

Explanations

references to screens or display-related concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_0/width_16k/average_l0_105

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.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

География

-0.57

 reservations

-0.57

(@"

-0.57

Портал

-0.56

:])

-0.55

 BoxFit

-0.53

 Meksiku

-0.53

 experimenta

-0.53

 umgekehrt

-0.52

zaw

-0.52

POSITIVE LOGITS

 screen

0.96

screen

0.94

 Belief

0.85

 screens

0.82

 SCREEN

0.81

creen

0.79

Belief

0.79

belief

0.78

 excuse

0.78

 Screens

0.75

Activations Density 0.169%

references to screens or display-related concepts

No Comments

No Known Activations