INDEX

Explanations

terms related to scientific study or classification

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-65k

Configuration

google/gemma-scope-2b-pt-res/layer_0/width_65k/average_l0_73

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

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

 cherchés

-0.66

ografija

-0.62

couvrez

-0.61

IndentedString

-0.61

withIdentifier

-0.57

ophysics

-0.56

couverte

-0.55

География

-0.54

__*/

-0.54

woners

-0.54

POSITIVE LOGITS

 Stud

1.05

 stud

0.92

Stud

0.89

Stan

0.81

 Flour

0.80

 ARTHUR

0.79

stan

0.78

 Stan

0.77

 stan

0.75

 Terminal

0.75

Activations Density 0.101%

terms related to scientific study or classification

No Comments

No Known Activations