INDEX

Explanations

notices where people are prompted to read more about a given topic

oai_token-act-pair · gpt-3.5-turbo

references to "Read More" prompts in a text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

Ń·

-0.76

ģĸ

-0.73

»Ĵ

-0.73

ranch

-0.64

phia

-0.63

 Rite

-0.62

sole

-0.61

irrel

-0.61

 Cree

-0.60

EStreamFrame

-0.60

POSITIVE LOGITS

 VIDEOS

1.06

 SHARES

0.70

 NEWS

0.69

 MORE

0.68

 Detected

0.66

ING

0.66

agascar

0.66

 INFORMATION

0.66

 ARTICLE

0.65

 STOR

0.64

Activations Density 0.022%

notices where people are prompted to read more about a given topic

references to "Read More" prompts in a text

No Comments

No Known Activations