INDEX

Explanations

words related to political and economic issues

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

bender

-0.78

puff

-0.71

cum

-0.70

tar

-0.69

icter

-0.66

conom

-0.66

wic

-0.65

wrap

-0.64

ussen

-0.64

more

-0.64

POSITIVE LOGITS

selves

1.38

own

1.20

 ancestors

1.02

 beloved

1.00

 selves

0.95

 ourselves

0.93

 asses

0.93

 adversaries

0.92

 collective

0.90

 hearts

0.88

Activations Density 0.315%

words related to political and economic issues

No Comments

No Known Activations