INDEX

Explanations

phrases indicating combinations or mixtures of elements

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

-0.17

icc

-0.15

ush

-0.15

orate

-0.14

åĹ

-0.14

 Sebast

-0.14

urous

-0.14

Ã±a

-0.13

 åı

-0.13

POSITIVE LOGITS

ÑħÐ¾Ð´

0.16

 circumstances

0.15

 sexes

0.15

ä¸ľè¥¿

0.14

edReader

0.14

 Ð·Ð°ÑģÑĤ

0.14

chine

0.14

Fixture

0.13

 elements

0.13

 Mort

0.13

Activations Density 0.026%

phrases indicating combinations or mixtures of elements

No Comments

No Known Activations