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

ships

-0.74

opot

-0.74

agar

-0.69

abus

-0.68

nah

-0.68

 ener

-0.67

entin

-0.66

ogl

-0.66

espie

-0.65

osponsors

-0.64

POSITIVE LOGITS

gb

0.75

ipolar

0.75

ACE

0.75

aternity

0.73

XL

0.72

 Clause

0.70

displayText

0.68

 Duel

0.68

©¶æ

0.65

 Parenthood

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.