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

yton

-0.75

XM

-0.72

enhagen

-0.70

ciating

-0.69

Compare

-0.66

Rail

-0.63

ourt

-0.63

ebook

-0.62

javascript

-0.62

Football

-0.62

POSITIVE LOGITS

ĳ

0.93

Ķ

0.87

Ľ

0.77

İĭ

0.76

st

0.73

ĺ

0.69

 Cheong

0.67

ī

0.66

¯

0.65

sto

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.