INDEX

Explanations

numbers and numerical values in various formats

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

 Bentley

-0.19

.eclipse

-0.16

 Rust

-0.14

BOVE

-0.14

 Bobby

-0.14

 Rosenberg

-0.14

 Wednesday

-0.14

ãģĹãĤĩãģĨ

-0.14

celed

-0.14

urgy

-0.14

POSITIVE LOGITS

Fri

0.38

 Friday

0.34

fri

0.34

Fri

0.32

 friday

0.32

Friday

0.31

 Fridays

0.29

0.28

0.26

0.25

Activations Density 0.077%

numbers and numerical values in various formats

No Comments

No Known Activations