INDEX

Explanations

words related to critical analysis, evaluation, or judgement

oai_token-act-pair · gpt-3.5-turbo

letters and abbreviations that could indicate various classifications or categorizations

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

©¶æ¥µ

-0.86

å§«

-0.78

LEY

-0.75

EStream

-0.72

ĳå£«

-0.72

 Shooter

-0.72

Cot

-0.69

 Lauder

-0.69

 Gunn

-0.68

 Lans

-0.66

POSITIVE LOGITS

actory

1.16

umbers

1.10

ynamic

1.04

elcome

0.99

arc

0.98

uzz

0.97

aint

0.97

ickets

0.97

esides

0.96

unning

0.96

Activations Density 0.111%

words related to critical analysis, evaluation, or judgement

letters and abbreviations that could indicate various classifications or categorizations

No Comments

No Known Activations