INDEX

Explanations

text related to the concept of improvement or enhancing something

oai_token-act-pair · gpt-3.5-turbo

phrases related to self-improvement and enhancement processes

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

cium

-0.86

DragonMagazine

-0.78

NetMessage

-0.77

cia

-0.69

ãĤ¦ãĤ¹

-0.68

kers

-0.68

runners

-0.66

mad

-0.66

py

-0.65

worth

-0.64

POSITIVE LOGITS

 efficiency

1.16

 responsiveness

0.98

 accuracy

0.97

 reliability

0.95

 quality

0.94

 hygiene

0.93

 outcomes

0.93

 upon

0.92

 coordination

0.92

 productivity

0.90

Activations Density 0.084%

text related to the concept of improvement or enhancing something

phrases related to self-improvement and enhancement processes

No Comments

No Known Activations