INDEX

Explanations

symbols and technical terms related to computing

oai_token-act-pair · gpt-3.5-turbo

punctuation and symbols often used in software documentation or issues

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

 lifes

-0.76

 unheard

-0.67

 concess

-0.67

 commission

-0.64

lamm

-0.64

 affili

-0.64

 legally

-0.64

 sick

-0.64

 credential

-0.64

 earthqu

-0.63

POSITIVE LOGITS

Conclusion

1.51

Lastly

1.42

Finally

1.37

CONCLUS

1.16

Secondly

1.14

Nevertheless

1.13

 Lastly

1.12

Overall

1.12

Furthermore

1.07

Alternatively

1.07

Activations Density 0.763%

symbols and technical terms related to computing

punctuation and symbols often used in software documentation or issues

No Comments

No Known Activations