INDEX

Explanations

the word "not"

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

ĸļ

-0.89

äºĶ

-0.83

¸

-0.81

comings

-0.78

SourceFile

-0.77

undrum

-0.73

nikov

-0.71

forts

-0.70

eway

-0.70

 srfAttach

-0.69

POSITIVE LOGITS

 they

0.93

 there

0.85

you

0.84

we

0.80

 someone

0.80

 qualifies

0.77

 anyone

0.76

he

0.73

 technically

0.73

 somebody

0.70

Activations Density 0.029%

the word "not"

No Comments

No Known Activations