INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

Disclaimer

-0.82

Whe

-0.76

GROUND

-0.71

nutrition

-0.70

Recipe

-0.68

Application

-0.66

 therein

-0.66

Nut

-0.65

ther

-0.65

site

-0.64

POSITIVE LOGITS

enf

0.71

selves

0.70

kered

0.70

Äĩ

0.68

BG

0.67

Ã±o

0.67

vard

0.66

ern

0.66

ibr

0.66

 ding

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.