INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 23-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L23R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.23.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

OfYear

-0.14

:animated

-0.14

vu

-0.14

isman

-0.14

çļĦä¸Ģä¸ª

-0.13

abet

-0.13

odate

-0.13

imp

-0.13

utherford

-0.12

 ëĦ¤ìĿ´íĬ¸

-0.12

POSITIVE LOGITS

 default

0.16

 following

0.16

ese

0.16

 behaviour

0.15

 second

0.15

 returned

0.15

 whole

0.15

 owning

0.15

 actual

0.15

 offending

0.14

Activations Density 0.346%

No Comments

No Known Activations