INDEX

Explanations

phrases and terms related to fighting against social issues like homelessness, poverty, and violence

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

ÅĻes

-0.16

 Invent

-0.15

rieving

-0.14

 ambigu

-0.14

å¹¸

-0.14

 pomoc

-0.14

ziej

-0.14

":[{↵

-0.14

 attacker

-0.14

 Niet

-0.14

POSITIVE LOGITS

 poverty

0.23

 hunger

0.21

 cycles

0.19

 homelessness

0.18

 inequality

0.18

 structural

0.18

 violence

0.18

 Poverty

0.18

 climate

0.18

 division

0.18

Activations Density 0.244%

phrases and terms related to fighting against social issues like homelessness, poverty, and violence

No Comments

No Known Activations