INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Â©

-0.76

laughs

-0.76

Eps

-0.73

raph

-0.73

 Paula

-0.70

cia

-0.67

mania

-0.67

 Hulu

-0.66

yright

-0.66

ogether

-0.65

POSITIVE LOGITS

gie

0.74

 breeds

0.67

ruciating

0.64

VERT

0.64

 stockpile

0.60

 nonviolent

0.60

 mutated

0.59

Rust

0.58

 variance

0.57

 Sharp

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.