INDEX

Explanations

references to authors and citations in academic work

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

anlÄ±k

-0.16

è¨Ģãģ£ãģŁ

-0.15

asaki

-0.15

enas

-0.14

-fw

-0.14

avor

-0.14

avors

-0.14

 Phong

-0.14

ouser

-0.14

oday

-0.14

POSITIVE LOGITS

di

0.25

 Rossi

0.21

 Grass

0.21

 Buzz

0.21

 Russo

0.20

 Giuliani

0.20

 Rover

0.20

 Frances

0.20

 Amend

0.20

 Vital

0.20

Activations Density 0.068%

references to authors and citations in academic work

No Comments

No Known Activations