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

tera

-0.82

mast

-0.79

âĵĺ

-0.77

ombies

-0.73

cv

-0.70

azeera

-0.68

rg

-0.68

xual

-0.68

mag

-0.65

 Nept

-0.65

POSITIVE LOGITS

 Trad

0.65

 Choi

0.61

urn

0.61

 Griff

0.61

 Kare

0.60

owed

0.60

Tem

0.59

 Jaguars

0.59

Rocket

0.58

 Ming

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.