INDEX

Explanations

titles indicating different scenarios or topics for discussion

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate varying perspectives or opinions

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

 conclusion

-0.66

 recap

-0.65

 attm

-0.62

 guiActiveUn

-0.61

 Deliver

-0.60

arov

-0.60

});

-0.59

VERTISEMENT

-0.59

obin

-0.58

 forward

-0.58

POSITIVE LOGITS

icion

0.84

orsi

0.72

pires

0.63

 dearly

0.62

 sake

0.61

pired

0.61

illusion

0.60

istors

0.60

cemic

0.59

chers

0.59

Activations Density 0.147%

titles indicating different scenarios or topics for discussion

phrases that indicate varying perspectives or opinions

No Comments

No Known Activations