INDEX

Explanations

phrases related to government policies and political statements

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

 slang

-0.63

OUT

-0.62

iard

-0.61

 accelerator

-0.61

Beginning

-0.60

Proxy

-0.60

Quant

-0.60

Mode

-0.59

 jargon

-0.59

 quizz

-0.58

POSITIVE LOGITS

 dignity

1.17

 wellbeing

1.11

 livelihood

1.11

 integrity

1.07

 humankind

1.01

 unborn

0.98

 humanity

0.95

 sake

0.94

 biodiversity

0.94

 grandchildren

0.94

Activations Density 1.801%

phrases related to government policies and political statements

No Comments

No Known Activations