INDEX

Explanations

phrases indicating completion or conclusion

oai_token-act-pair · gpt-3.5-turbo

phrases indicating the concept of something being "over."

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Deity

-0.67

 partName

-0.64

AMA

-0.62

osity

-0.62

yssey

-0.59

 Enhancement

-0.59

 Forward

-0.57

 ALSO

-0.56

estial

-0.55

iosity

-0.55

POSITIVE LOGITS

rated

1.26

kill

1.25

stated

1.19

blown

1.18

awed

1.12

valued

1.12

priced

1.10

reaching

1.09

loading

1.08

bearing

1.07

Activations Density 0.030%

phrases indicating completion or conclusion

phrases indicating the concept of something being "over."

No Comments

No Known Activations