INDEX

Explanations

references to government entities and their actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

 national

-0.27

 nationwide

-0.26

 Congressional

-0.24

 NATIONAL

-0.24

 congressional

-0.24

 Federal

-0.23

Federal

-0.22

 Ð½Ð°ÑĨÐ¸Ð¾Ð½Ð°Ð»ÑĮ

-0.21

national

-0.20

nation

-0.20

POSITIVE LOGITS

 statewide

0.49

 Governor

0.40

 state

0.36

 governor

0.36

 State

0.34

å·ŀ

0.33

state

0.31

 gubern

0.28

 Legislature

0.28

 province

0.27

Activations Density 0.188%

references to government entities and their actions

No Comments

No Known Activations