INDEX

Explanations

words related to regulations, bans, and warnings related to substances

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

issance

-0.88

swick

-0.86

sonian

-0.83

estine

-0.80

pocket

-0.78

irgin

-0.76

ilus

-0.74

esta

-0.73

venture

-0.73

apest

-0.72

POSITIVE LOGITS

 imposed

1.38

 enforced

1.17

 prohibiting

1.06

 levied

1.03

 forb

1.02

 violations

0.96

 restricting

0.94

 exemptions

0.93

 enforce

0.93

 compliance

0.92

Activations Density 3.127%

words related to regulations, bans, and warnings related to substances

No Comments

No Known Activations