INDEX

Explanations

words related to commands and directives

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

ahat

-0.16

gger

-0.14

 Bing

-0.14

ilen

-0.14

zim

-0.13

 ÐļÑĢÐ°

-0.13

imar

-0.13

 Ð¢ÐµÐ¼

-0.13

bern

-0.13

POSITIVE LOGITS

ekli

0.17

 Georgetown

0.15

È

0.15

uko

0.15

mts

0.15

izik

0.15

ãĥ¼ãĥī

0.15

Peb

0.14

emple

0.14

haul

0.14

Activations Density 0.006%

words related to commands and directives

No Comments

No Known Activations