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

 teasp

-0.80

tein

-0.76

 gunmen

-0.72

 Tanz

-0.71

uala

-0.71

ertodd

-0.68

 sacks

-0.68

avorite

-0.66

xual

-0.66

KT

-0.65

POSITIVE LOGITS

Â¯

0.77

 Riding

0.70

ãĤĬ

0.69

aic

0.68

Ģ

0.67

linger

0.66

 Casting

0.65

ãĤī

0.61

LOCK

0.61

wcsstore

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.