INDEX

Explanations

markers of excitement and anticipation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

iface

-0.15

rine

-0.14

angers

-0.14

Ø¬Ø±

-0.14

 Bowman

-0.13

 rarely

-0.13

idders

-0.13

 bazen

-0.13

essaging

-0.13

AndView

-0.13

POSITIVE LOGITS

 stay

0.42

 Stay

0.37

 fingers

0.36

Stay

0.36

stay

0.34

 stays

0.30

 stayed

0.29

 expect

0.27

 watch

0.26

 staying

0.25

Activations Density 0.169%

markers of excitement and anticipation

No Comments

No Known Activations