INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 10-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L10R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.10.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

ÅĽ

-0.16

ects

-0.15

echa

-0.15

acter

-0.14

æ°ĹæĮģãģ¡

-0.14

rouch

-0.14

urchase

-0.14

amp

-0.14

ÑĥÐºÑĤ

-0.14

lege

-0.14

POSITIVE LOGITS

0.14

 Elemental

0.14

asl

0.13

idon

0.13

 Lust

0.13

universal

0.13

reso

0.13

.kotlin

0.13

Activations Density 0.000%

No Known Activations

This feature has no known activations.