INDEX

Explanations

references to abstract concepts and moral themes, particularly related to charity and ethics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

 Kane

-0.30

 Wade

-0.29

rite

-0.28

inge

-0.28

lane

-0.28

oke

-0.28

 italiane

-0.28

lete

-0.28

pose

-0.27

pute

-0.27

POSITIVE LOGITS

lovak

0.14

esidir

0.13

CLUD

0.13

anism

0.13

acidad

0.13

zon

0.13

ison

0.13

anium

0.13

entin

0.12

 VerfÃ¼g

0.12

Activations Density 0.523%

references to abstract concepts and moral themes, particularly related to charity and ethics

No Comments

No Known Activations