INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

 seperate

-0.16

 recieved

-0.16

uries

-0.15

à¤¬à¤²

-0.15

æŁı

-0.15

chor

-0.14

efon

-0.13

ury

-0.13

epend

-0.13

POSITIVE LOGITS

/stdc

0.15

sal

0.15

inspace

0.14

owie

0.14

ThanOrEqualTo

0.13

aven

0.13

coln

0.13

 BITTE

0.13

-plus

0.13

-columns

0.13

Activations Density 0.000%

No Known Activations

This feature has no known activations.