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

 corrections

-0.81

ãĤ©

-0.72

 shortcuts

-0.70

 overlook

-0.66

 streak

-0.66

 haven

-0.65

«ĺ

-0.65

 icons

-0.65

 underestimate

-0.64

inx

-0.64

POSITIVE LOGITS

uctor

0.82

rice

0.77

Subject

0.77

rica

0.77

oglu

0.76

 Seller

0.73

eers

0.69

Fish

0.68

anza

0.68

ibia

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.