INDEX

Explanations

references to legal violations and breaches of laws or regulations

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_fs12288-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

12,288

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

hin

-0.73

iller

-0.65

erers

-0.64

rogens

-0.64

flush

-0.64

rians

-0.63

naissance

-0.63

inse

-0.62

uilding

-0.62

alk

-0.61

POSITIVE LOGITS

 laws

1.24

 statutes

1.15

 regulations

1.11

 Rule

1.11

 rules

1.11

 norms

1.10

 confidentiality

1.09

 Laws

1.08

 treaties

1.07

 contractual

1.05

Activations Density 0.379%

references to legal violations and breaches of laws or regulations

No Comments

No Known Activations