INDEX

Explanations

direct references to tech concepts or discussions

oai_token-act-pair · gpt-3.5-turbo

negative statements or contractions related to the word "isn't."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 shifts

-0.75

 Hats

-0.63

jog

-0.62

 exploits

-0.61

 blitz

-0.59

 defeats

-0.59

 Gaul

-0.58

 Doodle

-0.58

GP

-0.57

 flies

-0.57

POSITIVE LOGITS

0.98

ï¸ı

0.97

tal

0.93

exempt

0.89

ti

0.88

mental

0.86

ready

0.85

stellar

0.85

minent

0.85

achable

0.83

Activations Density 0.146%

direct references to tech concepts or discussions

negative statements or contractions related to the word "isn't."

No Comments

No Known Activations