INDEX

Explanations

comparative phrases and examples illustrating similarities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

-0.15

ears

-0.15

reau

-0.14

âĢİ

-0.14

orno

-0.14

buie

-0.13

ellen

-0.13

¹

-0.13

 CascadeType

-0.13

-append

-0.12

POSITIVE LOGITS

 ones

0.24

ones

0.22

 Ð½Ð°Ð¿ÑĢÐ¸Ð¼ÐµÑĢ

0.19

ebek

0.17

 napÅĻÃŃklad

0.16

ONES

0.16

 Ones

0.16

 ÙħØ«ÙĦØ§

0.16

Notifier

0.15

zeros

0.15

Activations Density 0.065%

comparative phrases and examples illustrating similarities

No Comments

No Known Activations