INDEX

Explanations

references to specific dangerous or harmful events

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ÏģÎ¸

-0.14

 bitwise

-0.13

imest

-0.13

ulp

-0.12

 Savings

-0.12

avern

-0.12

datable

-0.12

ìĤ´

-0.12

.setTo

-0.11

 analyzer

-0.11

POSITIVE LOGITS

 security

0.39

 Security

0.33

security

0.31

-security

0.29

 escorts

0.29

Security

0.29

 escorted

0.28

VIP

0.27

 SECURITY

0.27

 escort

0.27

Activations Density 0.299%

references to specific dangerous or harmful events

No Comments

No Known Activations