INDEX

Explanations

specific discussion points or topics referenced by <person>'s observations or work

oai_token-act-pair · gpt-3.5-turbo

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

Reviewer

-0.76

 Tradable

-0.74

oke

-0.72

eeee

-0.69

oldown

-0.68

icio

-0.67

atu

-0.66

hered

-0.65

 Skies

-0.65

Season

-0.65

POSITIVE LOGITS

 findings

1.38

 remarks

1.32

 conclusions

1.31

 assertion

1.30

 observations

1.28

 analysis

1.25

 assertions

1.23

 characterization

1.22

 hypothesis

1.22

 conclusion

1.22

Activations Density 0.353%

specific discussion points or topics referenced by <person>'s observations or work

No Comments

No Known Activations