INDEX

Explanations

references to various types of threats or dangers

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

Negative Logits

alink

-0.17

iao

-0.16

ureka

-0.16

quine

-0.15

uib

-0.14

è¢ĸ

-0.14

 ÑĢÐ°ÑģÑħÐ¾Ð´

-0.14

paque

-0.13

ecided

-0.13

 dÆ°á»¡ng

-0.13

POSITIVE LOGITS

 posed

0.42

posed

0.30

ening

0.29

ened

0.27

 level

0.24

 hanging

0.24

 looming

0.24

eners

0.23

Pos

0.23

 against

0.22

Activations Density 0.041%

references to various types of threats or dangers

No Known Activations