INDEX

Explanations

references to publications and manuscript processes

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

abestanden

-0.66

canActivate

-0.60

 circunstan

-0.60

DebuggerNonUser

-0.58

arschu

-0.56

retario

-0.55

ctor

-0.55

ngua

-0.55

()?;

-0.55

'&:

-0.54

POSITIVE LOGITS

 publications

0.74

 Publications

0.68

Fitting

0.63

 imagination

0.62

 communications

0.62

 PUBLICATION

0.62

 publication

0.61

epiece

0.61

 fitting

0.60

Publication

0.59

Activations Density 0.557%

references to publications and manuscript processes

No Comments

No Known Activations