INDEX

Explanations

phrases indicating the potential to improve or take action

oai_token-act-pair · gpt-3.5-turbo

phrases indicating the ability or possibility of action

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

mes

-0.64

Winged

-0.63

Listener

-0.62

IL

-0.62

ilty

-0.61

 Related

-0.59

reth

-0.59

Lenin

-0.58

idelines

-0.57

Cly

-0.56

POSITIVE LOGITS

 easily

1.11

 safely

1.01

 cheaply

0.92

 anywhere

0.90

 feas

0.87

 comfortably

0.86

 plaus

0.85

ivably

0.84

 anytime

0.81

 whatever

0.80

Activations Density 0.255%

phrases indicating the potential to improve or take action

phrases indicating the ability or possibility of action

No Comments

No Known Activations