INDEX

Explanations

numbers indicating quantity or measurement

oai_token-act-pair · gpt-3.5-turbo

numerical ratings and scores

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Canaver

-0.73

é¾į

-0.72

 conflic

-0.70

SPONSORED

-0.70

alties

-0.60

ersen

-0.59

 perspect

-0.59

sers

-0.58

 sensibilities

-0.57

taboola

-0.57

POSITIVE LOGITS

2.77

2.27

2.19

2.08

1.97

1.85

1.78

1.76

1.74

1.61

Activations Density 0.037%

numbers indicating quantity or measurement

numerical ratings and scores

No Comments

No Known Activations