INDEX

Explanations

news-related topics and current events, particularly negative incidents or controversies

oai_token-act-pair · gpt-3.5-turbo

phrases related to legal and ethical violations

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

;)

-0.78

tho

-0.72

!!

-0.71

:-)

-0.69

 Thoughts

-0.63

aughs

-0.62

:,

-0.61

:)

-0.59

}}}

-0.57

Fn

-0.56

POSITIVE LOGITS

 watchdog

0.62

 alleging

0.61

 illegally

0.59

 transgender

0.59

 allegedly

0.57

 detained

0.56

 unlawfully

0.56

 belonging

0.56

 underwater

0.55

 relating

0.55

Activations Density 1.262%

news-related topics and current events, particularly negative incidents or controversies

phrases related to legal and ethical violations

No Comments

No Known Activations