INDEX

Explanations

references to destruction or demolition events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 20-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L20R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.20.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

ariat

-0.16

-0.14

 æģ

-0.14

rail

-0.14

OLUM

-0.14

íĻĢ

-0.14

ØŃØ¯Ø©

-0.13

Ð¾ÑĨÑĸ

-0.13

 sealed

-0.13

POSITIVE LOGITS

 destruction

0.29

 destroyed

0.28

destroy

0.24

 destroy

0.24

 building

0.24

 Destruction

0.24

.destroy

0.24

 buildings

0.23

raz

0.23

 demolition

0.22

Activations Density 0.123%

references to destruction or demolition events

No Comments

No Known Activations