Serdar Ozsoy

50 posts

Serdar Ozsoy

Serdar Ozsoy

@ozsoyserdar

PhD student @UniBonn

Katılım Ocak 2013
168 Takip Edilen46 Takipçiler
Serdar Ozsoy retweetledi
Deniz Yuret
Deniz Yuret@denizyuret·
Have you ever seen a learning curve that looks like a step function? It turns out a few hundred negative examples flips a switch inside an LLM and gives a discrete jump in accuracy. "How much do LLMs learn from negative examples?" (arxiv.org/abs/2503.14391) with @ShadiSHamdan.
Deniz Yuret tweet media
English
0
1
15
2.3K
Serdar Ozsoy
Serdar Ozsoy@ozsoyserdar·
My new habit, starting with reasoning models: not going to the final answer without reading the chain of thought. Less efficient, more enjoyable.
English
0
0
0
60
Serdar Ozsoy
Serdar Ozsoy@ozsoyserdar·
The new content race on social media: DeepSeek.
English
0
0
0
95
Serdar Ozsoy retweetledi
Alper Erdogan
Alper Erdogan@Alper_T_E·
Join us at the poster session for our #Neurips2023 article "Correlative Information Maximization: A Biologically Plausible Approach to Supervised Deep Neural Networks without Weight Symmetry", our joint work with @BozkurtBariscan & @CPehlevan. 🗓️ Wed, 5-7 p.m. 📍 Hall B1+B2 #423
Alper Erdogan tweet media
English
2
4
34
4.2K
Serdar Ozsoy retweetledi
Bariscan Bozkurt
Bariscan Bozkurt@BozkurtBariscan·
🎉Thrilled to announce that our paper for #NeurIPS2023 titled “Correlative Information Maximization: A Biologically Plausible Approach to Supervised Deep Neural Networks without Weight Symmetry” has been accepted! (arxiv.org/abs/2306.04810)
Bariscan Bozkurt tweet media
English
1
9
63
9K
Serdar Ozsoy retweetledi
Alper Erdogan
Alper Erdogan@Alper_T_E·
Confused about #NeurIPS2023 rebuttal policy. Can we submit multiple threads for each reviewer? With 6000 characters limitation, it is hard to respond to all reviewer comments/questions. Can we also use "Official Comment" to provide our responses.
English
0
3
9
1.5K
Serdar Ozsoy retweetledi
KUIS AI
KUIS AI@KuisAICenter·
Here are the papers: "Biologically-Plausible Determinant Maximization Neural Networks for Blind Separation of Correlated Sources" #NeurIPS2022 (github.com/bariscanbozkur…)
English
1
1
5
0
Serdar Ozsoy retweetledi
Sercan Arık
Sercan Arık@sercanarik·
We’re excited to announce that TabNet is now available in Vertex AI Tabular Workflows: lnkd.in/g7jB5EMe! Tabular Workflows provides fully managed, optimized, and scalable pipelines, making it easier to use TabNet without worrying about implementation details. (1/3)
English
1
1
5
0
Serdar Ozsoy retweetledi
Alper Erdogan
Alper Erdogan@Alper_T_E·
The existing biologically plausible neural network approaches to solve the blind source separation problem typically assume independence/uncorrelatedness of sources. In our article, we propose an alternative framework that is capable of separating correlated sources.
Alper Erdogan tweet media
English
2
6
19
0
Serdar Ozsoy retweetledi
Deniz Yuret
Deniz Yuret@denizyuret·
Self supervised learning is revolutionizing AI using large unlabeled datasets. We show that maximizing mutual information between alternative representations of the same input is a practical method for self supervised learning that is immune to the dreaded collapse problem.
Deniz Yuret tweet media
English
1
46
344
0
Serdar Ozsoy
Serdar Ozsoy@ozsoyserdar·
In Pytorch, recommendation not to use bias=True for linear or conv layers before BatchNorm (BN) is not correct when using ReLU between them. Bias is useless since BN centers the values again but this is not the case when ReLU exists before BN. (Ex: Linear-ReLU-BN)
English
0
0
0
0
Serdar Ozsoy
Serdar Ozsoy@ozsoyserdar·
Oversampling should be done after test-train splitting and only for train split. This is also valid for cross-validation. Otherwise that causes 1)data leakage due to duplicated samples, 2)biased performance measure due to test-valid set difference. The same goes for SMOTE.
English
0
0
1
0
Serdar Ozsoy
Serdar Ozsoy@ozsoyserdar·
While people have been discussing ethical side of moral decision of driverless car, now doctors have to decide who will have treatment in case of limited resources in reality.
English
0
0
0
0
Serdar Ozsoy
Serdar Ozsoy@ozsoyserdar·
- Every matrix can be triangularized by choosing an orthonormal basis. - Really, are you "Schur"?
English
0
0
4
0

Keşfet

@ShadiSHamdan @BozkurtBariscan @CPehlevan @Alper_T_E @denizyuret @KuisAICenter @ftm_guney @akaan_akan