INDEX

Explanations

references to specific strategy games and their mechanics

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

lew

-0.15

 Ð´Ð²ÐµÑĢ

-0.15

asma

-0.15

æµ¦

-0.15

PFN

-0.15

ìķħ

-0.14

NFC

-0.14

itom

-0.14

ozor

-0.14

æŃ¦åĻ¨

-0.14

POSITIVE LOGITS

civ

0.29

 civil

0.28

 Civilization

0.28

Civ

0.28

 Civil

0.27

Civil

0.25

civil

0.25

åįķä½į

0.24

 units

0.24

 strategy

0.23

Activations Density 0.200%

references to specific strategy games and their mechanics

No Comments

No Known Activations