INDEX

Explanations

words related to reading the next part of a text or content

oai_token-act-pair · gpt-3.5-turbo

references to forms of content navigation or article links

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ĪĴ

-0.75

itialized

-0.67

ascist

-0.66

lease

-0.65

arse

-0.65

 simul

-0.65

ranch

-0.64

gone

-0.61

gran

-0.60

 clutch

-0.60

POSITIVE LOGITS

 WATCHED

1.18

 VIDEOS

0.94

 IMAGES

0.90

 Expand

0.89

 HuffPost

0.86

 Thumbnails

0.80

 LINK

0.79

 Loading

0.78

Featured

0.76

 âĢº

0.76

Activations Density 0.260%

words related to reading the next part of a text or content

references to forms of content navigation or article links

No Comments

No Known Activations