INDEX

Explanations

technical terms and concepts related to engine performance and upgrades

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

ount

-0.17

 chute

-0.15

iras

-0.15

ountains

-0.15

ÃŃda

-0.14

oref

-0.14

ichert

-0.14

aklÄ±

-0.14

æŁĵ

-0.13

ucci

-0.13

POSITIVE LOGITS

cam

0.29

 cams

0.25

 pist

0.24

 intake

0.23

 timing

0.22

 bore

0.22

 stroke

0.22

 combustion

0.22

 valves

0.21

 cylinder

0.21

Activations Density 0.022%

technical terms and concepts related to engine performance and upgrades

No Comments

No Known Activations