INDEX

Explanations

actions and commands related to user input and selection processes

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

isset

-0.19

ackers

-0.19

æĪĲäºĨ

-0.16

akers

-0.16

akens

-0.15

æĮģãģ¡

-0.15

åĪ°äºĨ

-0.15

ÙĨØ¯Ùĩ

-0.15

imeType

-0.14

izada

-0.14

POSITIVE LOGITS

owaÄĩ

0.31

ificar

0.31

Ð¸ÑĤÑĮ

0.30

Ð²Ð°ÑĤÑĮ

0.29

ar

0.28

Ð¾Ð²Ð°ÑĤÑĮ

0.28

sar

0.26

entar

0.25

strar

0.25

Ð½ÑĥÑĤÑĮ

0.24

Activations Density 0.031%

actions and commands related to user input and selection processes

No Comments

No Known Activations