INDEX

Explanations

specific keywords and terms related to various contexts such as programming, parameters, locations, and specific project elements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 10-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L10R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.10.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

oman

-0.15

ottes

-0.15

DOM

-0.14

ushi

-0.14

edik

-0.14

Us

-0.14

ilen

-0.13

sla

-0.13

 Parade

-0.13

POSITIVE LOGITS

Ð´Ð°Ð½

0.16

ecies

0.15

 tard

0.14

 itself

0.14

olon

0.14

 Tiffany

0.14

ibu

0.14

ound

0.13

)prepare

0.13

å¤ªéĥİ

0.13

Activations Density 0.272%

specific keywords and terms related to various contexts such as programming, parameters, locations, and specific project elements

No Comments

No Known Activations