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

bis

-0.17

 incel

-0.16

yb

-0.15

umas

-0.14

-vs

-0.14

(nullptr

-0.14

enschaft

-0.13

ãĢ

-0.13

ervo

-0.13

chant

-0.13

POSITIVE LOGITS

 Brid

0.21

 conference

0.19

 conferences

0.19

 Conference

0.18

idea

0.17

 Professional

0.16

 Blog

0.16

 Bridges

0.16

-twitter

0.16

PD

0.16

Activations Density 0.000%

No Known Activations

This feature has no known activations.