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

Duc

-0.74

theless

-0.70

orius

-0.65

"}],"

-0.65

rieg

-0.65

"],"

-0.64

esan

-0.64

 å¤

-0.63

lessly

-0.61

walking

-0.60

POSITIVE LOGITS

brance

0.76

phies

0.72

irmed

0.71

sted

0.67

urities

0.66

 Kerr

0.64

athom

0.64

ole

0.63

ticket

0.62

 passport

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.