INDEX

Explanations

specific unique identifiers or symbols often seen in academic or technical contexts

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

 authorised

-0.14

 Almighty

-0.13

 enjo

-0.13

.om

-0.13

ocommerce

-0.13

ĸī

-0.13

-END

-0.13

yll

-0.13

 mortal

-0.13

 SPELL

-0.12

POSITIVE LOGITS

 culturally

0.25

 Japanese

0.24

 cultural

0.24

 collect

0.23

 Cultural

0.22

 japanese

0.22

 western

0.21

 culture

0.21

Japanese

0.21

 contextual

0.20

Activations Density 0.003%

specific unique identifiers or symbols often seen in academic or technical contexts

No Comments

No Known Activations