INDEX

Explanations

phrases or sentences related to emphasizing a point or argument

oai_token-act-pair · gpt-3.5-turbo

quantitative descriptors preceding nouns or phrases

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

eyes

-0.81

breakers

-0.78

chairs

-0.77

ographs

-0.76

assisted

-0.75

anism

-0.74

ares

-0.72

friends

-0.69

stars

-0.69

OWS

-0.69

POSITIVE LOGITS

 tendency

1.28

 possibility

1.26

 difference

1.21

lot

1.18

 reason

1.16

 shortage

1.15

 discrepancy

1.12

 downside

1.12

 misconception

1.12

 temptation

1.11

Activations Density 0.102%

phrases or sentences related to emphasizing a point or argument

quantitative descriptors preceding nouns or phrases

No Comments

No Known Activations