INDEX

Explanations

terms related to education or learning

oai_token-act-pair · gpt-3.5-turbo

content related to education

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

lesh

-0.77

lane

-0.73

mask

-0.72

 Chains

-0.69

sha

-0.69

 Toro

-0.68

 Wraith

-0.68

pus

-0.67

wrap

-0.66

 Giants

-0.66

POSITIVE LOGITS

 educational

3.43

 Educational

2.50

 education

2.24

educ

2.21

 educ

2.12

Educ

2.12

education

2.01

Education

1.99

 educators

1.96

 instructional

1.91

Activations Density 0.025%

terms related to education or learning

content related to education

No Comments

No Known Activations