INDEX

Explanations

phrases related to design and building entities or organizations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 11-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L11R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.11.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

oji

-0.16

èĵ

-0.15

ccione

-0.15

Den

-0.14

 poster

-0.13

alette

-0.13

 Welfare

-0.13

 Goodman

-0.13

ollo

-0.13

annel

-0.13

POSITIVE LOGITS

estone

0.16

cha

0.15

amar

0.14

CHANT

0.14

 é¦

0.14

zy

0.14

PIN

0.13

kaar

0.13

 deut

0.13

 powers

0.13

Activations Density 0.025%

phrases related to design and building entities or organizations

No Comments

No Known Activations