INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

gp

-0.74

opian

-0.73

igl

-0.71

ours

-0.71

elaide

-0.70

ullivan

-0.70

 Constantine

-0.70

imp

-0.67

nc

-0.67

agascar

-0.67

POSITIVE LOGITS

lessness

0.89

less

0.76

 slic

0.66

enh

0.65

itar

0.63

 eval

0.60

 overload

0.60

 emoji

0.59

 change

0.59

 speed

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.