INDEX

Explanations

references to the act of reviewing or evaluating

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

ad

-0.17

cola

-0.16

fall

-0.15

way

-0.15

wa

-0.14

cor

-0.14

zego

-0.14

aze

-0.14

ARA

-0.14

strict

-0.13

POSITIVE LOGITS

 cref

0.17

OrCreate

0.16

ÏģÏīÎ½

0.15

anian

0.15

 ìĤ¬íķŃ

0.15

Î¬Î½Ïī

0.15

raÄį

0.15

tro

0.15

/comment

0.14

Ø§ÙĩÙħ

0.14

Activations Density 0.040%

references to the act of reviewing or evaluating

No Comments

No Known Activations