INDEX

Explanations

references to academic roles and research interests

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

ube

-0.17

 Mathematic

-0.16

erre

-0.15

chen

-0.15

 mathematic

-0.15

ÑĢÐ°Ñĩ

-0.15

Physics

-0.15

 Scientist

-0.14

 physicists

-0.14

 scientist

-0.14

POSITIVE LOGITS

 Area

0.25

 Critical

0.25

 critical

0.24

 area

0.24

 literature

0.23

 comparative

0.23

Area

0.22

Compar

0.21

Critical

0.21

 liter

0.21

Activations Density 0.174%

references to academic roles and research interests

No Comments

No Known Activations