INDEX

Explanations

phrases related to a question and answer format, potentially involving a unique identifier like 'Q'

oai_token-act-pair · gpt-3.5-turbo

interrogative phrases and structures

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

undai

-0.73

querque

-0.64

chnology

-0.61

 undermin

-0.59

ividual

-0.59

oulos

-0.59

conservancy

-0.58

 phosphate

-0.57

olicy

-0.55

 Thornton

-0.55

POSITIVE LOGITS

Boss

0.74

Trivia

0.74

âĶĢâĶĢâĶĢâĶĢ

0.72

________________

0.70

<@

0.70

talk

0.67

 ----------------------------------------------------------------

0.65

 Conclusion

0.64

--------------------------------------------------------

0.64

 =================================================================

0.63

Activations Density 0.679%

phrases related to a question and answer format, potentially involving a unique identifier like 'Q'

interrogative phrases and structures

No Comments

No Known Activations