INDEX

Explanations

complimentary expressions or positive remarks

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

conservancy

-0.79

-0.74

perature

-0.73

çķ

-0.72

ç«

-0.70

è¦ļéĨĴ

-0.68

rift

-0.67

ensis

-0.66

è

-0.66

uca

-0.65

POSITIVE LOGITS

ly

0.98

 Thoughts

0.92

 kidding

0.89

huh

0.87

bye

0.82

 Facts

0.79

 Advice

0.78

rontal

0.75

 Enough

0.75

Sounds

0.75

Activations Density 0.179%

complimentary expressions or positive remarks

No Comments

No Known Activations