INDEX

Explanations

symbols and characters that represent encoded text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

Paz

-0.17

 jewellery

-0.16

Æ¯

-0.14

 SHIPPING

-0.14

dac

-0.13

-Javadoc

-0.13

.jd

-0.13

Labour

-0.13

idden

-0.12

met

-0.12

POSITIVE LOGITS

 broadband

0.24

 municipal

0.23

 Mount

0.22

 ISPs

0.22

 municip

0.21

 Municipal

0.21

Mount

0.20

 Connectivity

0.20

 community

0.20

 telecommunications

0.20

Activations Density 0.003%

symbols and characters that represent encoded text

No Comments

No Known Activations