INDEX

Explanations

references to political movements, locations, and organizations

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_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

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

 channels

-0.88

 corridors

-0.82

 establishments

-0.81

 Machines

-0.80

NetMessage

-0.77

 embassies

-0.77

sites

-0.77

tics

-0.77

 classes

-0.77

 scenes

-0.77

POSITIVE LOGITS

 consisting

1.07

 resembling

1.06

 akin

1.03

 Called

0.97

 whose

0.96

 called

0.95

 reminiscent

0.94

 specializing

0.89

 wherein

0.87

 worthy

0.87

Activations Density 6.352%

references to political movements, locations, and organizations

No Comments

No Known Activations