INDEX

Explanations

references to events or competitions involving child models

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

NJ

-0.19

Nz

-0.18

NJ

-0.18

 Maduro

-0.17

Florida

-0.17

 Boca

-0.16

 Venezuel

-0.16

nj

-0.16

 Serbian

-0.16

velt

-0.15

POSITIVE LOGITS

 Leeds

0.62

 Yorkshire

0.46

 Bradford

0.46

 Wake

0.33

York

0.33

 Calder

0.32

 Kirk

0.32

 Halifax

0.31

 york

0.31

LS

0.30

Activations Density 0.021%

references to events or competitions involving child models

No Comments

No Known Activations