INDEX

Explanations

phrases related to political discussions and controversies

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

Negative Logits

sensitive

-0.70

bles

-0.70

pmwiki

-0.69

 spoilers

-0.69

ALLY

-0.68

obyl

-0.68

 parcel

-0.68

artifacts

-0.66

nesses

-0.66

 agre

-0.66

POSITIVE LOGITS

 antiquity

0.93

 course

0.83

 Europe

0.82

 Appalach

0.82

 legend

0.82

 Northeast

0.81

 Brill

0.80

WW

0.78

 adolescence

0.78

Activations Density 0.144%

phrases related to political discussions and controversies

No Known Activations