INDEX

Explanations

phrases related to providing instructions, specifically focusing on guiding someone through a process or giving directions on what to do

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

theless

-0.90

anwhile

-0.75

 Mous

-0.69

 DRAG

-0.65

ESA

-0.65

 Triangle

-0.64

 Newton

-0.62

 Springer

-0.61

 Thomson

-0.60

 dispers

-0.60

POSITIVE LOGITS

¤

1.46

ª

1.42

Ŀ

1.42

¡

1.40

¶

1.39

¥

1.37

¦

1.36

´

1.33

£

1.32

¢

1.32

Activations Density 0.127%

phrases related to providing instructions, specifically focusing on guiding someone through a process or giving directions on what to do

No Comments

No Known Activations