INDEX

Explanations

terms and concepts related to scholarly works and original content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 11-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L11R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.11.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

jang

-0.15

zell

-0.15

ise

-0.14

ç·

-0.14

uce

-0.14

angle

-0.14

ÑıÐº

-0.14

 éĩ

-0.14

ige

-0.14

ke

-0.13

POSITIVE LOGITS

anos

0.17

licit

0.16

wiÄħ

0.15

íħ

0.14

Graphic

0.14

defs

0.14

ertino

0.14

)const

0.14

.Atomic

0.14

Ø§ÙĪÛĮØ±

0.14

Activations Density 0.278%

terms and concepts related to scholarly works and original content

No Comments

No Known Activations