INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

EStreamFrame

-0.79

Forward

-0.70

kus

-0.69

 Tycoon

-0.66

 Pros

-0.64

Buy

-0.64

Sense

-0.64

leness

-0.64

hiba

-0.63

vre

-0.63

POSITIVE LOGITS

ctuary

1.01

bryce

0.73

BLIC

0.63

nah

0.62

hement

0.61

[_

0.61

uled

0.61

requ

0.60

oubt

0.60

issa

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.