INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

ely

-0.20

atik

-0.14

lius

-0.14

esis

-0.14

esh

-0.14

actly

-0.14

enton

-0.14

eli

-0.13

ively

-0.13

egin

-0.13

POSITIVE LOGITS

-navbar

0.15

igidBody

0.15

 Prem

0.14

redo

0.14

 Wich

0.14

ltre

0.14

 Bund

0.14

 prem

0.14

Attrib

0.13

quo

0.13

Activations Density 1.076%

No Comments

No Known Activations