INDEX

Explanations

locations within a building where specific actions or events took place

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

 incumb

-0.76

":["

-0.71

 importantly

-0.68

Reviewer

-0.67

 thous

-0.67

few

-0.65

llor

-0.65

 executes

-0.64

 constitu

-0.64

 disrupt

-0.63

POSITIVE LOGITS

 lieu

1.12

 Cannes

0.98

ked

0.96

0.95

 vitro

0.95

 conjunction

0.95

 accordance

0.95

patient

0.95

0.94

Activations Density 0.319%

locations within a building where specific actions or events took place

No Comments

No Known Activations