INDEX

Explanations

terms related to clothing, specifically focusing on pants

oai_token-act-pair · gpt-3.5-turbo

instances of the word "pantheon."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

UTION

-0.82

 Sett

-0.71

 Carbuncle

-0.71

wards

-0.70

ITED

-0.68

FORMATION

-0.66

KNOWN

-0.65

FINE

-0.65

upon

-0.65

ptives

-0.65

POSITIVE LOGITS

heon

1.38

 pant

1.22

agraph

0.97

icip

0.94

 Pant

0.91

hetically

0.86

ry

0.83

ĸļ

0.83

oleon

0.83

agonist

0.82

Activations Density 0.004%

terms related to clothing, specifically focusing on pants

instances of the word "pantheon."

No Comments

No Known Activations