INDEX

Explanations

references to academic journal issues and volumes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 16-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L16R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.16.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

 early

-0.20

 blind

-0.19

 earliest

-0.16

Early

-0.15

æĹ©

-0.15

åĪĿ

-0.15

chart

-0.14

 earlier

-0.14

Ãłng

-0.14

 prim

-0.14

POSITIVE LOGITS

ekil

0.15

ObjectContext

0.15

Wor

0.15

MaxY

0.15

arring

0.15

xmax

0.14

getLast

0.14

reative

0.14

entine

0.14

ÐµÐºÐ¾Ð¼

0.14

Activations Density 0.007%

references to academic journal issues and volumes

No Comments

No Known Activations