INDEX

Explanations

descriptive words and phrases related to entertainment and consumer products

oai_token-act-pair · gpt-3.5-turbo

different forms of entertainment such as movies, games, books, and music

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

 embassies

-0.77

 occupations

-0.76

NetMessage

-0.76

igham

-0.72

å§«

-0.70

aples

-0.64

 clauses

-0.64

iths

-0.63

 Cosponsors

-0.63

 chromosomes

-0.63

POSITIVE LOGITS

 featuring

0.91

 aimed

0.89

 combining

0.86

 nonetheless

0.85

 designed

0.83

 showcasing

0.82

 unlike

0.82

mable

0.80

 thats

0.78

 devoid

0.75

Activations Density 0.325%

descriptive words and phrases related to entertainment and consumer products

different forms of entertainment such as movies, games, books, and music

No Comments

No Known Activations