INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

ucks

-0.16

printStats

-0.16

dens

-0.15

cano

-0.15

eil

-0.15

iert

-0.15

artment

-0.15

URLException

-0.14

UILayout

-0.14

tro

-0.14

POSITIVE LOGITS

ALSE

0.15

ye

0.14

ayo

0.14

ÑģÐ¾ÑĢ

0.13

Tr

0.13

lfw

0.13

bose

0.13

nas

0.13

-equiv

0.13

Activations Density 0.000%

No Known Activations

This feature has no known activations.