INDEX

Explanations

personal information, names, and descriptive details

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.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

enegger

-0.65

ĸļ

-0.64

 afterward

-0.63

_.

-0.61

 Frie

-0.56

 afterwards

-0.56

 beforehand

-0.56

Peb

-0.55

[*

-0.55

ï¸

-0.55

POSITIVE LOGITS

tones

0.60

isations

0.56

coins

0.56

Â¶

0.55

occup

0.54

 cancers

0.52

ÂŃ

0.52

RNA

0.52

âĢĲ

0.52

isms

0.52

Activations Density 0.602%

personal information, names, and descriptive details

No Comments

No Known Activations