INDEX

Explanations

mentions of specific authorities, officials, police, or investigators

oai_token-act-pair · gpt-3.5-turbo

references to authorities or officials involved in various situations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

%%

-0.60

©¶æ

-0.60

 parallels

-0.59

 tread

-0.59

DragonMagazine

-0.56

paralle

-0.55

 pedest

-0.54

 analogy

-0.52

 Thumbnails

-0.52

~~

-0.51

POSITIVE LOGITS

 said

1.26

said

1.02

 told

0.99

 wrote

0.99

 reported

0.98

 announced

0.98

 says

0.95

 confirmed

0.91

 disclosed

0.91

 testified

0.89

Activations Density 0.274%

mentions of specific authorities, officials, police, or investigators

references to authorities or officials involved in various situations

No Comments

No Known Activations