INDEX

Explanations

numerical ranges or figures

oai_token-act-pair · gpt-3.5-turbo

numerical data related to measurements and quantities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

-0.80

tainment

-0.72

 Muse

-0.70

 Lauder

-0.64

EStream

-0.64

 Satan

-0.64

rolet

-0.63

Haunted

-0.62

atem

-0.61

Trend

-0.60

POSITIVE LOGITS

 minutes

0.87

 consecutive

0.85

 msec

0.83

 strokes

0.83

 percent

0.82

ecause

0.79

 seconds

0.79

cm

0.78

 cents

0.75

 trillion

0.74

Activations Density 0.315%

numerical ranges or figures

numerical data related to measurements and quantities

No Comments

No Known Activations