INDEX

Explanations

phrases that indicate quantities or collections

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 5-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L5R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.5.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

raud

-0.17

ëĳ

-0.15

/extensions

-0.15

Gig

-0.15

edge

-0.15

-0.14

urnished

-0.14

eed

-0.14

-0.13

ider

-0.13

POSITIVE LOGITS

ptron

0.18

aroo

0.16

Ð¸

0.15

Menus

0.15

 poru

0.14

 trunk

0.14

ButtonClick

0.14

ecast

0.14

strap

0.14

iest

0.14

Activations Density 0.013%

phrases that indicate quantities or collections

No Comments

No Known Activations