INDEX

Explanations

phrases starting with slash, possibly related to news articles or online posts

oai_token-act-pair · gpt-3.5-turbo

references to image sources or photo credits

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

 haun

-0.68

 bonding

-0.64

avis

-0.64

 blush

-0.60

 bond

-0.59

 resur

-0.59

 loose

-0.59

 debates

-0.58

 Wake

-0.58

 Deer

-0.58

POSITIVE LOGITS

Getty

1.63

GET

1.23

File

1.17

AFP

1.04

Reuters

0.98

 Getty

0.92

Images

0.92

Files

0.89

License

0.86

Zero

0.84

Activations Density 0.032%

phrases starting with slash, possibly related to news articles or online posts

references to image sources or photo credits

No Comments

No Known Activations