INDEX

Explanations

questions directed at an audience

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 10-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L10R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.10.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

ÑĮÑİ

-0.15

MIX

-0.14

gh

-0.14

sad

-0.14

asy

-0.14

antro

-0.14

usu

-0.14

eyen

-0.13

erts

-0.13

ilon

-0.13

POSITIVE LOGITS

ipa

0.15

 cÃ´ng

0.14

 Roller

0.14

atsapp

0.14

Yug

0.14

Serialization

0.13

 Vulcan

0.13

agal

0.13

&type

0.13

Jog

0.13

Activations Density 0.118%

questions directed at an audience

No Comments

No Known Activations