INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 28-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L28R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.28.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

owi

-0.15

ushima

-0.14

 Ð½Ð¾Ð³

-0.14

 ÑĢÐµÐ±ÐµÐ½Ð¾Ðº

-0.14

fir

-0.14

IRC

-0.13

aptive

-0.13

 aucun

-0.13

ÑĢÐµÐ±

-0.13

 Everything

-0.13

POSITIVE LOGITS

 guys

0.55

 Guys

0.42

gu

0.35

guy

0.35

Guy

0.33

Guy

0.32

 folks

0.31

-all

0.29

all

0.29

fol

0.28

Activations Density 0.123%

No Comments

No Known Activations