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

ombat

-0.14

é»İ

-0.14

AVC

-0.14

 Lore

-0.14

↵

-0.14

afen

-0.13

ëŀĺ

-0.13

 âĢİ

-0.13

Ð´ÐµÐ»

-0.13

,↵

-0.13

POSITIVE LOGITS

 generosity

0.20

 extrem

0.19

 Herbert

0.19

geber

0.17

 extreme

0.16

 skept

0.15

 Gener

0.15

 speaker

0.15

 canceled

0.14

felt

0.14

Activations Density 0.000%

No Known Activations

This feature has no known activations.