INDEX

Explanations

terms and phrases related to visual perception and sights

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

hr

-0.16

hell

-0.16

head

-0.15

headed

-0.15

ãĥ¼ãĤ¯

-0.15

uchen

-0.15

halb

-0.14

oui

-0.14

ucch

-0.14

Ã¤tze

-0.14

POSITIVE LOGITS

seeing

0.42

Seeing

0.22

lessly

0.22

 sight

0.21

 Sight

0.19

geist

0.19

seen

0.18

ting

0.18

edly

0.17

ed

0.17

Activations Density 0.012%

terms and phrases related to visual perception and sights

No Comments

No Known Activations