INDEX

Explanations

mentions of the word "ama"

oai_token-act-pair · gpt-3.5-turbo

references to the word "ama" and its variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

sheet

-0.81

tle

-0.78

ting

-0.76

working

-0.73

points

-0.70

 clipboard

-0.66

 boil

-0.65

lay

-0.64

ried

-0.62

lines

-0.61

POSITIVE LOGITS

ama

1.32

amas

0.98

qua

0.94

ican

0.92

uthor

0.90

icans

0.88

Ã±

0.88

agi

0.88

Ã±a

0.87

ikan

0.85

Activations Density 0.005%

mentions of the word "ama"

references to the word "ama" and its variations

No Comments

No Known Activations