INDEX

Explanations

names of individuals involved in leadership positions

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

aleigh

-0.17

 æ®

-0.17

 Durham

-0.16

 phosph

-0.16

awan

-0.15

 Assad

-0.15

 Hawaiian

-0.15

 Kohana

-0.15

 ApplicationException

-0.15

 ÎĶÎ·

-0.14

POSITIVE LOGITS

 Gujarat

0.36

 àª

0.31

àª

0.29

 Gujar

0.29

 Ahmed

0.27

Vad

0.23

à«

0.21

iben

0.20

 Patel

0.20

gu

0.20

Activations Density 0.044%

names of individuals involved in leadership positions

No Comments

No Known Activations