INDEX

Explanations

references to historical events and cultural shifts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

Ã²i

-0.19

 Steele

-0.17

aukee

-0.16

ushima

-0.15

borg

-0.15

åĢ

-0.14

mscorlib

-0.14

éĴ¢

-0.14

igers

-0.14

ä¹ĥ

-0.14

POSITIVE LOGITS

 hipp

0.45

 Hipp

0.37

0.35

LSD

0.33

 Vietnam

0.33

hip

0.31

 Beat

0.30

 count

0.29

ipp

0.29

 Acid

0.29

Activations Density 0.309%

references to historical events and cultural shifts

No Comments

No Known Activations