INDEX

Explanations

phrases related to controversial topics or arguments

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

yards

-0.71

 seams

-0.66

 swell

-0.64

 Thumbnails

-0.61

 Thrones

-0.60

Cod

-0.60

 Bridgewater

-0.58

phrine

-0.58

 conscience

-0.57

 responsibly

-0.57

POSITIVE LOGITS

azard

1.00

ouston

0.99

awk

0.97

ype

0.97

ospital

0.97

ouse

0.96

ollow

0.95

rss

0.94

onen

0.93

uana

0.93

Activations Density 0.019%

phrases related to controversial topics or arguments

No Comments

No Known Activations