INDEX

Explanations

references to specific locations and geographical features

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Cuer

-0.57

 joan

-0.54

COUVER

-0.52

 Sarm

-0.52

 crud

-0.52

 Muses

-0.52

зулта

-0.52

 strix

-0.52

 bullion

-0.51

 tfsi

-0.51

POSITIVE LOGITS

AndEndTag

0.67

ny

0.66

 Madagascar

0.65

tsy

0.64

Madagascar

0.59

Ny

0.57

Ant

0.56

fy

0.54

Ny

0.53

ary

0.53

Activations Density 0.076%

references to specific locations and geographical features

No Comments

No Known Activations