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

 VIDEOS

-0.84

 transitions

-0.63

otypes

-0.60

dropping

-0.60

======

-0.59

 substitution

-0.58

inges

-0.57

eport

-0.56

 quitting

-0.56

 mistakes

-0.56

POSITIVE LOGITS

wit

0.80

noon

0.71

held

0.68

aday

0.67

wed

0.66

voc

0.62

vs

0.62

 Magicka

0.62

Roaming

0.61

WE

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.