INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 depended

-0.71

employment

-0.70

anchez

-0.67

 Nadu

-0.65

 deportation

-0.62

sections

-0.62

vere

-0.62

aunders

-0.60

 Mehran

-0.60

eenth

-0.60

POSITIVE LOGITS

OPA

0.73

æ©Ł

0.69

KT

0.66

 Cecil

0.66

PN

0.66

undai

0.63

TextColor

0.63

spl

0.63

Reviewer

0.62

 Scarlet

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.