INDEX

Explanations

phrases related to security threats and dangers

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

 Manufact

-0.70

endars

-0.68

£ı

-0.68

cknowled

-0.67

essen

-0.67

ciples

-0.64

 Colleges

-0.62

mith

-0.61

hent

-0.61

lang

-0.61

POSITIVE LOGITS

 hazard

1.28

 threat

1.26

 danger

1.21

 risk

1.16

 risks

1.12

 dangers

1.08

 hazards

1.06

 challenge

1.05

 peril

1.03

 jeopardy

1.00

Activations Density 0.125%

phrases related to security threats and dangers

No Comments

No Known Activations