arXiv subjects

Burak Ozdemir

Publications and source records attributed to Burak Ozdemir.

3 recordsLinked to original sources

arXiv2026

Universal Pathologies, Conditional Consequences: A Triple-Robustness Analysis of RAG for Multi-Hop Traceability

GraphRAG underperforms vector RAG on citation precision in many reports, but where and why have remained corpus-bound. We present a triple-robustness analysis that holds the retrieval architecture fixed and varies three orthogonal axes embedder (local e5-small -> Azure text-embedding-3-small), corpus (DO-178C typed-edge requirements -> Wikipedia paragraph chains via MuSiQue), and judge (paired GPT-5.4 x GPT-4.1) across 4,440 main-matrix runs, 600 cross-corpus runs, and 1,200 paired faithfulness judgments. (C2a) Over-citation is architecturally universal: GraphRAG emits 11-15 IDs per answer at citation precision 0.12-0.23 and retrieval recall 0.68-0.87 across all three settings. (C2b) Its faithfulness consequence is corpus-conditional: in typed-edge DO-178C, GraphRAG faithfulness collapses 74%->40% across hops; on Wikipedia chains the same pipeline rises 42%->58% because over-cited paragraphs remain topically supporting. (C1) Stratum-conditional winners are corpus-conditional but embedder-robust: vanilla wins 2-hop on DO-178C, GraphRAG wins 2-hop on MuSiQue, identical under either embedder. (C3) Single-judge LLM faithfulness is fragile to retrieval state: same-judge self-kappa across embedders is 0.137 for GPT-5.4 (verdict change on 41% of items). A learned router on dense embeddings alone reaches macro-F1 0.86 on hop classification (C4). We argue triple-robustness is the minimum bar for trustworthy RAG architecture claims.

cs.CL↗

arXiv2024

Comparison of Solar Cell Efficiencies of Black Phosphorus and Silicon at the Nano and Micro Scales from First-Principles Calculations

Density functional theory and many-body (GW+BSE) calculations of transmittance, absorbance, and reflectance are performed on silicon and black phosphorus (BP). We find that a damping value of 0.01 used in the dielectric function calculation is the optimal for calculating the solar cell efficiency of Si. Our calculations indicate that the solar cell efficiency of a 100 μm thick Si slab is 27.4% while the efficiency of BP for the same thickness is 2.33% and 1.94% for light polarized along the zigzag and armchair directions, respectively. For 100 nm thick materials, we obtain that Si presents a 0.8% solar cell efficiency and BP exhibits a 0.14% and 1.02% efficiency for light polarized along the the zigzag and armchair directions, respectively, indicating that BP performs better than Si at these small scales. Our results underscore the important effect of the material thickness on solar cell efficiencies.

cond-mat.mtrl-sci↗

arXiv2020

Black Phosphorus and Phosphorene/Graphene Heterostructure as Alkali metal (Li, Na, and K) Ion Battery

Black phosphorous is a layered material having a high capacity of 2596 mAh/g as a battery electrode, however it suffers from cracking due to high volume expansion during lithiation. These cracks causes loss of electrical contact in the whole material, therefore capacity fades after further cycles of charging and discharging. One needs a support material which would not crack with lithiation of phosphorous in order to keep the electrical contact of the material. Here, we considered phosphorene sandwiched between graphene layers. By using density functional theory, we calculated voltages of lithiation, sodiation, and potasiation of black phosphorous and phosphorene-graphene heterostructure which compares well with the experimental results. We found low voltages for both black phosphorous and phosphorene-graphene heterostructure therefore these materials can be used as an anode electrode in lithium-ion, sodium-ion, and potassium-ion batteries.

cond-mat.mtrl-sci↗