INDEX

Explanations

references to international meetings or conferences related to communication technology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

Pax

-0.15

eren

-0.15

blo

-0.15

iliz

-0.14

 Guinness

-0.14

ny

-0.14

Î³Ïĩ

-0.14

ump

-0.14

arra

-0.13

tul

-0.13

POSITIVE LOGITS

 Prep

0.24

 Ministers

0.21

 minister

0.21

 Minister

0.20

 ministers

0.19

 inters

0.18

 Bureau

0.17

 prepar

0.17

pl

0.17

Outcome

0.17

Activations Density 0.035%

references to international meetings or conferences related to communication technology

No Comments

No Known Activations