INDEX

Explanations

elements and tags in HTML code

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

 Ellis

-0.15

AGMA

-0.15

åĶ¯

-0.15

ÏĦÏģÎ¹

-0.14

#End

-0.14

afort

-0.14

Î¼Î¿Î½

-0.14

ANCH

-0.14

à¤ªà¤¤

-0.14

DetailsService

-0.14

POSITIVE LOGITS

<head

0.32

 head

0.31

head

0.30

 HEAD

0.25

 Head

0.25

Head

0.25

-head

0.23

/head

0.23

HEAD

0.22

.head

0.20

Activations Density 0.011%

elements and tags in HTML code

No Comments

No Known Activations