INDEX

Explanations

technical keywords and actions related to programming or coding

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

jectives

-0.14

lez

-0.14

asc

-0.14

Ì§

-0.13

 ConsoleColor

-0.12

à¸ªà¸²à¸§

-0.12

iat

-0.12

amed

-0.12

wan

-0.12

fty

-0.12

POSITIVE LOGITS

 something

0.38

:↵

0.37

 like

0.36

 thus

0.31

:↵↵

0.31

something

0.31

0.28

 Something

0.28

Something

0.28

å¦Ĥä¸ĭ

0.27

Activations Density 0.157%

technical keywords and actions related to programming or coding

No Comments

No Known Activations