INDEX

Explanations

references to articles and news reporting

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

chy

-0.17

ÑĢÐ¸Ð´

-0.16

PIO

-0.14

 Bomb

-0.14

arella

-0.14

Ð½Ð°ÑĢ

-0.14

ÑĩÐ¸Ð½Ð¸

-0.13

unh

-0.13

 Brend

-0.13

kiem

-0.13

POSITIVE LOGITS

 appeared

0.22

 originally

0.20

 appear

0.20

 appears

0.19

 appearance

0.19

Appear

0.18

 appearances

0.17

 apare

0.17

 Appears

0.16

appearance

0.16

Activations Density 0.028%

references to articles and news reporting

No Comments

No Known Activations