INDEX

Explanations

elements related to specific numerical data and coding terminology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

á½±

-0.29

ÙİÙĳ

-0.28

===↵

-0.27

á½·

-0.26

/*↵

-0.25

 /*↵

-0.25

á½³

-0.25

 percent

-0.24

===

-0.22

###

-0.21

POSITIVE LOGITS

 Â·

0.75

Â·

0.53

.hpp

0.52

 ####

0.49

####

0.45

_HPP

0.41

Â·Â·

0.34

 ÂłÂł

0.30

 ====

0.30

Â·Â·Â·Â·

0.30

Activations Density 0.049%

elements related to specific numerical data and coding terminology

No Comments

No Known Activations