INDEX

Explanations

references to government surveillance and watchlists

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 17-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L17R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.17.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

osaur

-0.14

achen

-0.14

ÑģÑĤÑĢ

-0.14

ambient

-0.14

imestamp

-0.14

 ä¹

-0.13

 mandate

-0.13

istine

-0.13

 Transparency

-0.13

rina

-0.13

POSITIVE LOGITS

 intelligence

0.21

 terrorism

0.21

 Intelligence

0.20

CVE

0.19

 Terrorism

0.19

 counter

0.17

al

0.17

CVE

0.16

AQ

0.16

 terrorist

0.16

Activations Density 0.133%

references to government surveillance and watchlists

No Comments

No Known Activations