INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

AFX

-0.15

Ð½Ð¾Ðº

-0.14

ulings

-0.14

å»ºç«ĭ

-0.14

ONUS

-0.14

olars

-0.13

*>*

-0.13

 Tomb

-0.13

rades

-0.13

 forfe

-0.13

POSITIVE LOGITS

drop

0.18

 research

0.17

drive

0.17

 researcher

0.17

 drop

0.17

research

0.16

 drive

0.16

Eval

0.16

 police

0.16

 respect

0.16

Activations Density 0.000%

No Known Activations

This feature has no known activations.