INDEX

Explanations

references to war and conflict, particularly related to Israel and Lebanon

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

 Uzbek

-0.20

 Zambia

-0.19

 Gujarat

-0.19

poÄįet

-0.18

 Birmingham

-0.18

-0.17

Alabama

-0.17

 Alabama

-0.17

 Rajasthan

-0.16

 bson

-0.16

POSITIVE LOGITS

 Lebanon

0.47

 Lebanese

0.46

 Beirut

0.46

 Leban

0.39

 Hezbollah

0.36

zbollah

0.27

 Cedar

0.26

Bek

0.26

 ÙĦØ¨

0.24

é»İ

0.24

Activations Density 0.040%

references to war and conflict, particularly related to Israel and Lebanon

No Comments

No Known Activations