INDEX

Explanations

quantifiable familial relationships and references to children or grandchildren

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

äºĮäºº

-0.20

iag

-0.17

-valu

-0.17

ä¸¤äºº

-0.15

iÄįka

-0.15

emplates

-0.15

éĺħè¯»æ¬¡æķ°

-0.15

ä¸Ģåį·

-0.14

hus

-0.14

Utf

-0.14

POSITIVE LOGITS

 sons

0.27

 children

0.24

 daughters

0.24

 boys

0.24

 grown

0.21

 Sons

0.21

 biological

0.20

 girls

0.19

 Children

0.19

 healthy

0.19

Activations Density 0.021%

quantifiable familial relationships and references to children or grandchildren

No Comments

No Known Activations