INDEX

Explanations

references to academic articles and their citations

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

elage

-0.21

awan

-0.16

egas

-0.15

Â°}

-0.15

 ransom

-0.14

erus

-0.14

aye

-0.14

Ä°R

-0.14

liches

-0.14

-0.13

POSITIVE LOGITS

 Google

0.30

Google

0.29

 google

0.24

.Google

0.23

_google

0.21

oogle

0.21

/google

0.21

-google

0.20

google

0.20

.google

0.20

Activations Density 0.091%

references to academic articles and their citations

No Comments

No Known Activations