INDEX

Explanations

social media-related terms including word counts

oai_token-act-pair · gpt-3.5-turbo

references to social media sharing activity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

tis

-0.78

arcity

-0.75

stress

-0.72

quartered

-0.72

 Frie

-0.70

smanship

-0.70

istically

-0.69

omsday

-0.65

urally

-0.64

seq

-0.63

POSITIVE LOGITS

 SHARES

1.05

 VIDEOS

0.98

 VIEW

0.96

 Facebook

0.90

 Share

0.88

Pinterest

0.86

 Email

0.86

LinkedIn

0.81

facebook

0.81

 facebook

0.80

Activations Density 0.020%

social media-related terms including word counts

references to social media sharing activity

No Comments

No Known Activations