INDEX

Explanations

phrases related to intentional actions

oai_token-act-pair · gpt-3.5-turbo

references to actions or behaviors, particularly those categorized as "acts."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Carlo

-0.72

ciating

-0.70

 Parad

-0.68

grown

-0.67

 Flavoring

-0.66

fortable

-0.66

 Galile

-0.65

 Keen

-0.64

lake

-0.61

 Dise

-0.60

POSITIVE LOGITS

uary

1.27

uated

1.20

ional

1.14

ivism

1.12

ivity

1.06

uate

1.05

uating

1.03

uation

0.99

inic

0.96

ives

0.93

Activations Density 0.019%

phrases related to intentional actions

references to actions or behaviors, particularly those categorized as "acts."

No Comments

No Known Activations