INDEX

Explanations

mentioning of specific numbers and topics such as movies or locations

oai_token-act-pair · gpt-3.5-turbo

references to photo captions and their associated context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 exponent

-0.84

 accordingly

-0.82

ively

-0.76

 proposition

-0.76

brainer

-0.74

 notwithstanding

-0.74

ministic

-0.73

pse

-0.73

instead

-0.73

 parity

-0.72

POSITIVE LOGITS

 Scenes

1.60

 Photos

1.56

 Images

1.20

 Moments

1.19

 Timeline

1.18

Photos

1.10

 Celeb

1.10

 IMAGES

1.10

 Faces

1.09

 Pictures

1.06

Activations Density 0.072%

mentioning of specific numbers and topics such as movies or locations

references to photo captions and their associated context

No Comments

No Known Activations