INDEX

Explanations

references to political oppression and dissent

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

 rebel

-0.16

 revolt

-0.15

holm

-0.15

 multer

-0.15

 rebels

-0.15

 cuckold

-0.14

adio

-0.14

dyby

-0.14

 inset

-0.14

 Rebellion

-0.14

POSITIVE LOGITS

 repression

0.33

 Crack

0.32

 clamp

0.31

 crackdown

0.31

 crack

0.29

 arbitrary

0.28

 cracking

0.28

 cracked

0.27

sil

0.26

 Clamp

0.26

Activations Density 0.270%

references to political oppression and dissent

No Comments

No Known Activations