INDEX

Explanations

words related to having a positive opinion or preference for something

oai_token-act-pair · gpt-3.5-turbo

expressions of approval or affection towards various subjects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

arta

-1.00

inas

-0.92

Dispatch

-0.81

pat

-0.75

por

-0.74

cession

-0.73

continental

-0.73

BT

-0.73

src

-0.73

chin

-0.72

POSITIVE LOGITS

lihood

1.67

 liking

1.11

 liked

0.96

0.92

liest

0.89

lier

0.87

 minded

0.86

 favorably

0.82

liness

0.79

 Likes

0.76

Activations Density 0.012%

words related to having a positive opinion or preference for something

expressions of approval or affection towards various subjects

No Comments

No Known Activations