INDEX

Explanations

concepts related to morality and decision-making

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

-0.77

 thous

-0.68

 hurd

-0.64

anwhile

-0.63

ikarp

-0.59

nces

-0.59

ãĥ¼ãĥĨãĤ£

-0.58

culosis

-0.55

vice

-0.55

mud

-0.55

POSITIVE LOGITS

 alike

1.22

 depending

1.17

depending

1.06

 respectively

0.95

 modes

0.71

 eras

0.71

0.69

0.66

BW

0.66

 dich

0.62

Activations Density 0.342%

concepts related to morality and decision-making

No Comments

No Known Activations