INDEX

Explanations

references to social interactions and community support

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

 wherever

-0.15

 Graf

-0.15

shopping

-0.14

=logging

-0.14

æĲº

-0.14

 account

-0.14

arra

-0.14

wind

-0.14

 wreak

-0.13

æ£®

-0.13

POSITIVE LOGITS

 hosting

0.25

 Hosting

0.23

 host

0.22

host

0.20

.host

0.20

 hosts

0.19

Hosting

0.19

/host

0.19

 welcoming

0.17

 hosted

0.17

Activations Density 0.158%

references to social interactions and community support

No Comments

No Known Activations