INDEX

Explanations

references to the "Ray-Knight theorem."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 19-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L19R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.19.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

Ã©ri

-0.15

loff

-0.13

 Hanging

-0.13

 Äĳai

-0.13

ãĥ³ãĥĢ

-0.13

ourg

-0.13

Å©

-0.13

itler

-0.13

uju

-0.13

Äįka

-0.13

POSITIVE LOGITS

Jay

0.21

Jay

0.19

ray

0.18

lay

0.17

Ray

0.17

avanaugh

0.16

RAY

0.16

Ray

0.16

Coy

0.16

LAY

0.16

Activations Density 0.040%

references to the "Ray-Knight theorem."

No Comments

No Known Activations