INDEX

Explanations

references to the character Iron Man or related terminology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.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

stral

-0.15

ourn

-0.15

inders

-0.15

iva

-0.14

omor

-0.14

á»Ĩ

-0.14

liers

-0.14

ÐµÑĩ

-0.14

ionario

-0.14

ibox

-0.14

POSITIVE LOGITS

ically

0.27

mong

0.26

ore

0.25

workers

0.23

iron

0.22

 Maiden

0.21

cl

0.21

 Iron

0.20

 maiden

0.20

Iron

0.20

Activations Density 0.009%

references to the character Iron Man or related terminology

No Comments

No Known Activations