INDEX

Explanations

social media metrics such as followers, views, likes, and comments

oai_token-act-pair · gpt-3.5-turbo

references to social media metrics and popularity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 manslaughter

-0.75

fug

-0.73

 suitcase

-0.73

 negotiators

-0.69

©¶æ

-0.69

 Quart

-0.68

EH

-0.65

iHUD

-0.64

Alz

-0.64

 Kirby

-0.64

POSITIVE LOGITS

Tweet

1.06

votes

1.06

hip

1.03

 Likes

0.99

ername

0.95

 retweet

0.93

hips

0.92

 Login

0.87

 followers

0.86

Posts

0.85

Activations Density 0.163%

social media metrics such as followers, views, likes, and comments

references to social media metrics and popularity

No Comments

No Known Activations