INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 28-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L28R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.28.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

udu

-0.16

ollo

-0.15

aan

-0.15

hell

-0.15

-0.14

ovÃ¡n

-0.14

ailer

-0.14

ugs

-0.14

ico

-0.14

shan

-0.14

POSITIVE LOGITS

rine

0.22

inka

0.21

leen

0.20

mand

0.20

rina

0.18

anning

0.18

MAND

0.17

rink

0.15

zen

0.15

Activations Density 0.009%

No Comments

No Known Activations