INDEX

Explanations

verb phrases related to hanging or being in a state of suspense

oai_token-act-pair · gpt-3.5-turbo

phrases related to socializing and spending time with others

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

ibel

-0.74

eger

-0.66

cean

-0.62

zed

-0.59

BE

-0.58

eele

-0.58

ledge

-0.57

anne

-0.57

tsky

-0.57

issan

-0.56

POSITIVE LOGITS

rily

1.16

overs

1.08

over

1.06

zhou

0.98

 upside

0.95

rier

0.95

 onto

0.94

outs

0.86

OVER

0.86

out

0.80

Activations Density 0.029%

verb phrases related to hanging or being in a state of suspense

phrases related to socializing and spending time with others

No Comments

No Known Activations