INDEX

Explanations

references to political appointments and affiliations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 22-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L22R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.22.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

thinkable

-0.15

isay

-0.15

 ÄĮer

-0.15

AFX

-0.14

Vys

-0.14

 Ryder

-0.14

 Bauer

-0.14

jte

-0.14

eyle

-0.14

utz

-0.14

POSITIVE LOGITS

 under

0.37

 Obama

0.35

 Barack

0.30

 Bush

0.30

Obama

0.29

under

0.28

 Trump

0.27

 Clinton

0.27

_under

0.25

 dÆ°á»Ľi

0.24

Activations Density 0.271%

references to political appointments and affiliations

No Comments

No Known Activations