INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

edded

-0.66

 dated

-0.60

 Mahm

-0.60

 Bangl

-0.60

 Assy

-0.59

 lasted

-0.59

sie

-0.58

Western

-0.58

 onward

-0.57

 Okin

-0.57

POSITIVE LOGITS

busters

0.71

TPPStreamerBot

0.70

isoft

0.68

 PACs

0.66

wcs

0.64

Leaks

0.63

 ransomware

0.63

Senate

0.63

 Ghostbusters

0.63

 charities

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.