INDEX

Explanations

instances of authorship or submission attribution in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

ÄĽn

-0.16

abeth

-0.15

 Habit

-0.15

ilder

-0.14

abis

-0.14

stick

-0.14

incer

-0.14

çĻ

-0.14

 Families

-0.13

inee

-0.13

POSITIVE LOGITS

Emm

0.17

avl

0.14

IRTH

0.14

ÏĦÎ¿ÏħÏĤ

0.14

 onto

0.14

 Ludwig

0.14

slice

0.14

egg

0.13

neck

0.13

ë¬¸íĻĶ

0.13

Activations Density 0.011%

instances of authorship or submission attribution in the text

No Comments

No Known Activations