INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

cps

-0.15

Ð¼Ð¸Ð½Ð¸

-0.15

ileo

-0.14

orsi

-0.14

 Morr

-0.14

 Morrison

-0.14

 dane

-0.14

ighb

-0.13

Bol

-0.13

vider

-0.13

POSITIVE LOGITS

 Harry

0.32

 Harold

0.29

Harry

0.29

 mods

0.28

 Rolex

0.28

Se

0.28

 watches

0.28

 watch

0.27

mod

0.26

Mods

0.25

Activations Density 0.000%

No Known Activations

This feature has no known activations.