INDEX

Explanations

titles and roles associated with authority figures and leadership positions

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

æģ

-0.14

ë²

-0.14

ãĤĵãģ©

-0.14

mpar

-0.14

 BITTE

-0.13

NST

-0.13

/change

-0.13

 itself

-0.13

ÑīÐ¾

-0.13

ALLEL

-0.13

POSITIVE LOGITS

-elect

0.27

 Emer

0.25

-President

0.20

ial

0.20

-design

0.20

-E

0.19

ess

0.18

-sama

0.17

/D

0.17

-General

0.17

Activations Density 0.133%

titles and roles associated with authority figures and leadership positions

No Comments

No Known Activations