INDEX

Explanations

various numerical and citation formatting elements commonly found in academic articles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

ãĤıãģĳ

-0.15

oucher

-0.15

esta

-0.14

essel

-0.14

.Formatting

-0.13

Ã©rÃ©

-0.13

posal

-0.13

venes

-0.13

azel

-0.13

POSITIVE LOGITS

ovit

0.14

eyJ

0.14

.ribbon

0.14

 Giles

0.14

ellig

0.13

.scalablytyped

0.13

elah

0.13

 preceded

0.13

Â£Â£

0.13

ÐµÐ½Ñģ

0.13

Activations Density 0.043%

various numerical and citation formatting elements commonly found in academic articles

No Comments

No Known Activations