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

thodox

-0.84

orneys

-0.84

monds

-0.80

lems

-0.79

otom

-0.76

Äĩ

-0.75

anish

-0.72

lÃ©

-0.72

itect

-0.71

eret

-0.71

POSITIVE LOGITS

Tik

0.80

 laureate

0.78

 Poker

0.75

 Knock

0.75

 Puzzle

0.74

 Wonderland

0.73

AU

0.72

 Mercenary

0.71

 Wanted

0.70

 Rampage

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.