INDEX

Explanations

phrases that imply questions about legality and correctness in processes

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

atron

-0.19

acha

-0.16

we

-0.15

aper

-0.15

 afterward

-0.14

 flavors

-0.14

 License

-0.14

èĶ

-0.14

*)_

-0.14

ä¿º

-0.13

POSITIVE LOGITS

 derog

0.16

xffffffff

0.16

 Keller

0.14

íİ¸

0.14

 Advocate

0.14

 Member

0.14

razy

0.14

 nationals

0.14

.raises

0.14

uder

0.14

Activations Density 0.002%

phrases that imply questions about legality and correctness in processes

No Comments

No Known Activations