INDEX

Explanations

words and phrases related to violence and death

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

Damage

-0.15

ä»ĭ

-0.15

amage

-0.15

ivec

-0.14

 Damage

-0.14

ÑĢÐµÑī

-0.14

damage

-0.13

ActionCreators

-0.13

esser

-0.13

urum

-0.13

POSITIVE LOGITS

 while

0.35

 whilst

0.31

 near

0.30

 outside

0.29

while

0.29

 during

0.27

 trying

0.27

 attempting

0.26

 WHILE

0.24

en

0.24

Activations Density 0.168%

words and phrases related to violence and death

No Comments

No Known Activations