INDEX

Explanations

phrases indicating guides, tips, or advice-oriented content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 23-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L23R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.23.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

Intialized

-0.17

WriteBarrier

-0.15

CloseOperation

-0.13

ãģ¬

-0.13

çŃ

-0.13

-With

-0.12

obao

-0.12

CallCheck

-0.12

nul

-0.12

ãĥ«ãĥķ

-0.12

POSITIVE LOGITS

 tips

0.38

how

0.37

 Tips

0.34

tips

0.32

How

0.32

 ways

0.31

how

0.30

Tips

0.29

why

0.28

How

0.28

Activations Density 0.232%

phrases indicating guides, tips, or advice-oriented content

No Comments

No Known Activations