INDEX

Explanations

temporal references to years and decades, highlighting the duration or timeline of events and experiences

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

 Seconds

-0.15

 zase

-0.14

 dÃ¼ÅŁ

-0.14

 Uncle

-0.14

 proceed

-0.14

ahl

-0.14

 unlimited

-0.14

ÐµÐ½Ð·

-0.14

itas

-0.14

æĹ¢çĦ¶

-0.14

POSITIVE LOGITS

 longer

0.28

 longest

0.25

 Longer

0.23

 finally

0.22

 during

0.21

æľŁéĹ´

0.21

during

0.21

 During

0.21

During

0.19

 durante

0.19

Activations Density 0.170%

temporal references to years and decades, highlighting the duration or timeline of events and experiences

No Comments

No Known Activations