INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L1R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.1.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

ergus

-0.17

ÑĢÐµÐ¹

-0.15

 Trev

-0.14

reib

-0.14

 toler

-0.14

Sel

-0.14

.mvp

-0.14

Ivy

-0.13

Sel

-0.13

Accept

-0.13

POSITIVE LOGITS

lator

0.15

baar

0.14

inker

0.14

iaux

0.14

Äįka

0.14

isay

0.14

Aur

0.14

Ð¾ÑĤÐ¸

0.14

 rabbits

0.13

edere

0.13

Activations Density 0.000%

No Known Activations

This feature has no known activations.