INDEX

Explanations

recommendations and instructions related to proper actions or guidelines

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

 Encounter

-0.15

 obliged

-0.15

 é¡

-0.14

à¸²à¸¢

-0.13

Äĥn

-0.13

kish

-0.13

/wiki

-0.13

Î²Î¿

-0.13

andler

-0.13

metro

-0.13

POSITIVE LOGITS

 remember

0.25

 familiar

0.24

 always

0.23

remember

0.21

 budget

0.21

budget

0.21

 never

0.20

 factor

0.20

 thoroughly

0.19

 Factor

0.19

Activations Density 0.297%

recommendations and instructions related to proper actions or guidelines

No Comments

No Known Activations