INDEX

Explanations

explicit statements or opinions

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

 Cruiser

-0.78

ibal

-0.67

swick

-0.63

 fingert

-0.61

ãĥł

-0.60

ãĥĥãĥĪ

-0.59

 Dwell

-0.56

ipel

-0.56

 faculties

-0.56

 Liver

-0.56

POSITIVE LOGITS

 aloud

1.34

 loudly

1.22

 goodbye

1.09

 louder

1.04

 loud

1.00

 bluff

0.90

 publicly

0.82

 farewell

0.81

 sarcast

0.79

 Goodbye

0.77

Activations Density 0.219%

explicit statements or opinions

No Comments

No Known Activations

explicit statements or opinions

No Comments

No Known Activations