INDEX

Explanations

phrases related to creating or causing something

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_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

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

thia

-0.67

phrine

-0.64

CLOSE

-0.63

assis

-0.60

Mania

-0.60

 Niet

-0.59

Fram

-0.58

heter

-0.56

ban

-0.55

pac

-0.55

POSITIVE LOGITS

 sure

1.20

hift

1.04

 sense

0.89

 landfall

0.89

 strides

0.86

 excuses

0.84

URE

0.82

 adjustments

0.82

 Sense

0.81

 headlines

0.79

Activations Density 2.135%

phrases related to creating or causing something

No Comments

No Known Activations