INDEX

Explanations

comparative phrases indicating degrees of magnitude or quantity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

æĽ´å¤ļ

-0.16

 Ð±Ð¾Ð»ÐµÐµ

-0.16

udem

-0.15

 MORE

-0.15

Ã¼gen

-0.15

 lebih

-0.15

 more

-0.15

Forge

-0.14

[~,

-0.14

POSITE

-0.14

POSITIVE LOGITS

 likely

0.31

 usual

0.29

 ever

0.28

likely

0.27

 anything

0.26

 meets

0.25

 anyone

0.25

 twice

0.24

ever

0.23

 anybody

0.22

Activations Density 0.040%

comparative phrases indicating degrees of magnitude or quantity

No Comments

No Known Activations