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

 homosexuality

-0.76

ulence

-0.76

 behavi

-0.74

gery

-0.74

_-

-0.68

 harmless

-0.68

frequency

-0.66

leness

-0.66

uton

-0.66

 behav

-0.66

POSITIVE LOGITS

sth

0.76

dds

0.66

 Dare

0.65

 Reps

0.65

FTWARE

0.63

aukee

0.63

 Crystal

0.63

ibilities

0.62

å§

0.61

eda

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.