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

Negative Logits

ilateral

-0.75

emale

-0.74

oggles

-0.74

emetery

-0.74

ilage

-0.69

RAY

-0.67

IRO

-0.67

rase

-0.66

uden

-0.66

owl

-0.63

POSITIVE LOGITS

 "$:/

0.86

 âĢº

0.71

othal

0.67

 Distribution

0.65

{{

0.65

"},{"

0.63

',

0.63

netflix

0.62

Cod

0.60

 earners

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.