INDEX

Explanations

references to annual events or milestones

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

kia

-0.15

)did

-0.14

annes

-0.14

Ø§Ø´Øª

-0.14

_Two

-0.14

 mÃª

-0.14

ulia

-0.14

criptors

-0.14

Ð¯Ðº

-0.13

tright

-0.13

POSITIVE LOGITS

 fifth

0.38

 fourth

0.35

 third

0.33

 sixth

0.33

 seventh

0.31

third

0.28

 eighth

0.27

 THIRD

0.27

ç¬¬åĽĽ

0.27

 ninth

0.26

Activations Density 0.073%

references to annual events or milestones

No Comments

No Known Activations