INDEX

Explanations

elements related to showcasing talent and performance improvement

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

 zend

-0.17

]={↵

-0.17

ignet

-0.16

ê°ĳ

-0.16

mtree

-0.15

apel

-0.15

ÑģÑĤÐ²Ð¾ÑĢ

-0.15

uhl

-0.15

ello

-0.14

ChangedEventArgs

-0.14

POSITIVE LOGITS

ester

0.15

redo

0.15

uada

0.15

san

0.14

 easily

0.14

Painter

0.14

vip

0.13

obre

0.13

 minim

0.13

_traits

0.13

Activations Density 0.397%

elements related to showcasing talent and performance improvement

No Comments

No Known Activations