INDEX

Explanations

references to performance or capacity in specific venues or contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

orage

-0.17

 Rowe

-0.16

ounder

-0.16

ÃŃas

-0.15

inski

-0.15

unate

-0.15

Deg

-0.15

 Atkins

-0.14

 Ames

-0.14

stantiate

-0.14

POSITIVE LOGITS

ium

0.53

rium

0.34

lum

0.32

mium

0.31

à¤¿à¤¯à¤®

0.31

yum

0.30

inium

0.29

gium

0.29

ulum

0.28

odium

0.28

Activations Density 0.083%

references to performance or capacity in specific venues or contexts

No Comments

No Known Activations