INDEX

Explanations

mentions of the word "captured"

oai_token-act-pair · gpt-3.5-turbo

references to specific scientific terms or phenomena related to illumination

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 guiIcon

-0.94

risome

-0.87

)]

-0.78

bernatorial

-0.74

¬¼

-0.73

stood

-0.73

itary

-0.71

mares

-0.71

isance

-0.70

ngth

-0.68

POSITIVE LOGITS

 millenn

0.65

 subconscious

0.62

OA

0.62

clair

0.61

 younger

0.60

 subt

0.60

 unsolved

0.59

 flipping

0.58

 tape

0.57

 dial

0.57

Activations Density 0.000%

mentions of the word "captured"

references to specific scientific terms or phenomena related to illumination

No Comments

No Known Activations