INDEX

Explanations

references to legal proceedings and trials

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

 arrests

-0.16

 Arrest

-0.16

ocide

-0.14

dish

-0.14

 noct

-0.13

 pÅĻij

-0.13

å¼¥

-0.12

 å¿

-0.12

Ãĭ

-0.12

 patrols

-0.12

POSITIVE LOGITS

 hearing

0.40

 hearings

0.32

 heard

0.31

 Hearing

0.30

 scheduled

0.29

 hear

0.27

adj

0.27

 session

0.26

 proceedings

0.26

 recess

0.26

Activations Density 0.088%

references to legal proceedings and trials

No Comments

No Known Activations