INDEX

Explanations

people or entities mentioned in news articles, especially politicians or public figures

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

taboola

-0.71

wagen

-0.65

 juggling

-0.61

 incremental

-0.61

muc

-0.61

 inconven

-0.60

 sshd

-0.60

 metaphor

-0.60

 contraction

-0.59

 levers

-0.58

POSITIVE LOGITS

aline

0.99

andise

0.80

itness

0.78

onso

0.77

thood

0.76

aida

0.75

bia

0.75

anto

0.73

asio

0.73

minus

0.73

Activations Density 0.045%

people or entities mentioned in news articles, especially politicians or public figures

No Comments

No Known Activations