INDEX

Explanations

personal pronouns followed by modal verbs or auxiliary verbs, indicating actions or intentions

oai_token-act-pair · gpt-3.5-turbo

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

 Gale

-0.67

groupon

-0.66

 Amid

-0.63

 Georgian

-0.63

advertising

-0.62

 passers

-0.61

 srfAttach

-0.61

 Rolls

-0.60

 Kenobi

-0.59

netflix

-0.58

POSITIVE LOGITS

'm

1.23

've

1.06

am

1.03

'll

1.01

 recommend

0.99

 propose

0.98

 intend

0.94

 apologize

0.94

 suggest

0.94

 apologise

0.94

Activations Density 0.138%

personal pronouns followed by modal verbs or auxiliary verbs, indicating actions or intentions

No Comments

No Known Activations