INDEX

Explanations

numerical values and formatting related to coding or styling

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

ë²Ķ

-0.17

 seiz

-0.17

APPER

-0.15

 plat

-0.15

raman

-0.14

iquer

-0.14

 MatSnackBar

-0.14

ossa

-0.13

 Huck

-0.13

ucher

-0.13

POSITIVE LOGITS

ipp

0.16

quential

0.16

sky

0.16

 çķ

0.15

ico

0.15

ICO

0.14

ÐµÐ»ÑĮ

0.14

izard

0.14

 Donovan

0.14

enso

0.13

Activations Density 0.016%

numerical values and formatting related to coding or styling

No Comments

No Known Activations