INDEX

Explanations

No Explanations Found

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

é²

-0.15

 Gins

-0.14

Pig

-0.13

obra

-0.13

altar

-0.13

 lurking

-0.13

ERV

-0.13

åĨ

-0.13

boa

-0.13

 Schwe

-0.13

POSITIVE LOGITS

dry

0.22

Dry

0.20

Dry

0.18

 berry

0.17

dry

0.17

 sprites

0.16

 sprite

0.16

 Papa

0.15

Sprites

0.15

prites

0.15

Activations Density 0.000%

No Known Activations

This feature has no known activations.