INDEX

Explanations

healthcare-related words and names

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

 Canaver

-0.54

product

-0.52

 namely

-0.51

 Defin

-0.51

 acutely

-0.50

 Dangerous

-0.50

é¾į

-0.49

ffer

-0.48

astically

-0.46

DEN

-0.46

POSITIVE LOGITS

etc

1.49

etc

0.91

0.82

,...

0.82

0.81

and

0.79

!,

0.72

+,

0.72

ect

0.71

?,

0.71

Activations Density 5.195%

healthcare-related words and names

No Comments

No Known Activations