INDEX

Explanations

numerical figures

oai_token-act-pair · gpt-3.5-turbo

statistics or numerical data associated with various events or circumstances

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

 causation

-0.67

 caus

-0.63

 awhile

-0.62

gue

-0.61

hetti

-0.59

itsch

-0.58

feed

-0.57

RET

-0.57

 haun

-0.56

zu

-0.56

POSITIVE LOGITS

 crammed

0.84

 Including

0.82

including

0.82

 apiece

0.82

 including

0.80

 totaling

0.79

 total

0.76

 collectively

0.74

 acres

0.73

illion

0.71

Activations Density 0.833%

numerical figures

statistics or numerical data associated with various events or circumstances

No Comments

No Known Activations