INDEX

Explanations

references to specific types of cobblestone or related words

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

zek

-0.18

udes

-0.17

 integral

-0.16

 Integral

-0.15

ISIBLE

-0.15

ermo

-0.15

æ³ķäºº

-0.14

integral

-0.14

 çĤ¹

-0.14

POSITIVE LOGITS

alt

0.39

webs

0.36

web

0.33

ble

0.33

ALT

0.29

blers

0.28

ertura

0.25

alto

0.25

WEB

0.23

-web

0.23

Activations Density 0.008%

references to specific types of cobblestone or related words

No Comments

No Known Activations