INDEX

Explanations

instances of the word "kom" or its variations, indicating a focus on components or parts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

bane

-0.19

ãĥ©ãĥ³ãĤ¹

-0.17

eft

-0.16

ãģŁãģĦ

-0.15

brahim

-0.15

eu

-0.15

 Lomb

-0.15

eper

-0.14

baugh

-0.14

yw

-0.14

POSITIVE LOGITS

pet

0.25

mand

0.25

pon

0.24

miss

0.24

mando

0.23

ite

0.22

pan

0.22

plex

0.21

odo

0.20

plet

0.20

Activations Density 0.004%

instances of the word "kom" or its variations, indicating a focus on components or parts

No Comments

No Known Activations