INDEX

Explanations

patterns and references related to technology and programming

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 28-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L28R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.28.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

 ØµØ¯

-0.16

enda

-0.15

ÑĸÐ½Ð½Ñı

-0.15

enz

-0.15

cape

-0.14

 mart

-0.14

bows

-0.14

ayan

-0.14

noinspection

-0.14

POSITIVE LOGITS

ano

0.28

anos

0.27

ani

0.25

ane

0.25

Ð°ÑĤÑĮ

0.23

ania

0.23

anie

0.23

Ð°Ð½Ð¾

0.23

ana

0.23

ali

0.22

Activations Density 0.028%

patterns and references related to technology and programming

No Comments

No Known Activations