INDEX

Explanations

text related to architecture and design

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

 defences

-0.73

 withdrawals

-0.70

Synopsis

-0.69

 rejection

-0.69

accept

-0.69

 propositions

-0.67

 judgments

-0.66

 defenses

-0.66

 compromises

-0.64

 isolation

-0.64

POSITIVE LOGITS

 Wikimedia

1.01

 Shutterstock

0.96

 Flickr

0.93

 STATS

0.85

 Estate

0.83

 Architects

0.81

 Everett

0.80

 IMAGES

0.79

 Courtesy

0.78

Edu

0.76

Activations Density 0.054%

text related to architecture and design

No Comments

No Known Activations