INDEX

Explanations

action verbs related to data manipulation and processing

oai_token-act-pair · gpt-3.5-turbo

verbs associated with data processing and organization

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

humans

-0.70

Enough

-0.58

endon

-0.56

 Americas

-0.54

¥µ

-0.54

 Shack

-0.53

 Britons

-0.53

 Sierra

-0.52

Tang

-0.52

 Blacks

-0.52

POSITIVE LOGITS

 separately

1.20

 individually

1.14

 automatically

1.12

 periodically

1.05

 manually

1.00

 randomly

1.00

 promptly

0.99

 anonymously

0.99

 accordingly

0.97

 electronically

0.96

Activations Density 0.357%

action verbs related to data manipulation and processing

verbs associated with data processing and organization

No Comments

No Known Activations