跳到正文
研踪PaperTrail

YOUR RESEARCH, IN SIGHT

让下一篇重要论文,不再错过。

覆盖计算机全学科,以 AI 为快捷入口。关注方向、作者和关键词,把发现变成持续的研究积累。

部分采集异常
正在读取账号与研究记录…

全站更新概况

+1072新增论文
+845其中 AI 相关
+94已有论文修订
+100新增 AI 导读

arXiv 最近完成扫描:2026-11-10 01:54 · 全站最近入库:2026-11-09 13:22各信源进度 →

统计口径与数据新鲜度

统计 2026-11-09 02:30 — 2026-11-10 02:30(北京时间) · 点击数字查看对应简报;AI 相关包含在新增论文内。

最近完成检查:2026-11-10 01:54最近发现新增版本:2026-11-09 13:22库内论文最新更新时间:2026-11-07 02:59

每分钟刷新已保存的数据。空检查不代表采集异常;新收录可能包含历史论文。AI 导读按新增版本计数。

18,172 已收录论文15,506 AI 相关4,527 有修订112 HF 社区发现

浏览论文库

时间范围跟随排序:版本更新、本站收录或首次发表。最新收录可能包含历史论文。部分信源异常,查看状态 →

找到 3 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-10 01:52(北京时间)
版本更新 · v4arXiv:2610.00329

Non-Commutative Lie-Algebraic State-Space Models: Finite-Division Variety Laws, Semidirect Cascades, and Gradient Saddle Escape

Zeyu Jia

原文摘要

Structured state-space models (SSMs) achieve linear-time inference and logarithmic-time parallel training via associative scans, yet homogeneous diagonal transitions generate solvable semigroups unable to track non-solvable group words or 3D rigid-body manifolds. We develop a theoretical and empirical framework for Non-Commutative Lie-Algebraic State-Space Models (Lie-SSMs), parameterizing recurrent updates on Lie algebra tangent spaces via the matrix exponential. First, via an o-minimal multi-rate cylindrical flag collar, we establish an unconditional single-layer finite-division criterion for homogeneous contracting $SO(3)$ and $SU(2)$ heads, proving odd-prime Sylow abelianity, odd-$p$-core centrality $[u^2,x]=e$, exponent-30 commutator laws ($[u^{30},v^{30}]=e$ in $SO(3)^m$ and $[u^{30},v^{30}]\in Z(G),[u^{30},v^{30}]^2=e$ in $SU(2)^m,SO(4)^m$, excluding $S_5$), and sharp threshold $d^*(S_5)=d^*(A_6)=5$ in $SO(5)$. Second, Depth-2 semidirect and Schreier $2$-cocycle cascades resolve $S_5$, $2I\cong SL_2(F_5)$, and $\mathrm{Heis}(F_p)$ in $O(\log T)$ depth, whereas unnormalized contracting coupling collapses exponentially without Directional RMSNorm. Third, we characterize the $O(\varepsilon^{c-1})$ Chen--Magnus--Hall--Witt commutator saddle and $1/|[G,G]|$ coset-centroid dilution, proving $\varepsilon=0$ symmetry breaking and local Polyak--Łojasiewicz convergence ($θ_{\max}<π$), solving $S_5$ ($99.90\%/99.40\%$ at $T=128$) and $D_{16}$ ($100.00\%$) via dense prefix loss and factored readouts. Across $Q_8,A_5,S_5,D_{16}$, Dyck-2, $SE_2(3)$ Left-Jacobian odometry, GPU scans ($26.03\times$ speedup), and three 3D IMU cohorts (UCI HAR, HAPT-PT 30-subject CV, and REALDISP), Depth-2 Lie-$SO(3)$ cascades reach $95.69\%\pm 0.14\%$ on HAR, $87.16\%\pm 0.64\%$ ($88.32\%\pm 0.93\%$ 30-subject episodes) on HAPT-PT, and $84.23\%\pm 1.00\%$ under displacement on REALDISP.

cs.LG

版本更新 · v2arXiv:2610.05900

Tri-Directional Sub-Voxel Displacement and Ipsilateral Territory Decoupling for 3D Bilateral Coronary Artery Tree Reconstruction in Cardiac CT Angiography

Zeyu Jia, Xin Ming

原文摘要

Three-dimensional (3D) reconstruction of bilateral coronary trees (LCA and RCA) from cardiac computed tomography angiography (CCTA) is limited by $64^3$ quantization, territory ambiguity, and sinotubular junction (STJ) ostium drift. We present a canonical $SO(3)$ $(Z_{\text{IS}}, Y_{\text{PA}}, X_{\text{LR}})$ multi-scale representation and Riemannian arborescence framework. First, we map 3D CCTA volumes, centerlines, and thick-stem ostia into $(Z_{\text{IS}}, Y_{\text{PA}}, X_{\text{LR}})$ space, extending a 3D pyramid and tubular decoder with a $128^3 \to 64^3$ boundary skip, a stop-gradient $128^3$ residual branch, a sub-voxel displacement head $Δ\hat{\mathbf{u}}(\mathbf{x})$, a dual-channel ipsilateral head $(\mathbf{P}_{\text{LCA}}, \mathbf{P}_{\text{RCA}})$, and a $32^3$ ostium detector. Second, all $327,385$ active parameters ($398,843$ total) are trained in one 30-epoch AdamW loop ($ε= 10^{-8}$) under decoupled L2-SP and aortic-wall consistency. Third, our Riemannian solver couples cylindrical Valsalva-STJ calibration, $128^3$ orifice contact rooting, contralateral inhibition, centripetal pruning, and 4-stage curvature-adaptive normal-plane pullback. On 14 unseen ImageCAS development volumes (28 trees; cases 201--250 isolated), the framework achieves $2.1652 \pm 0.4686\text{ mm}$ $\text{CD}_{\text{CL}}$ ($\text{CD}_{\text{CL}}^{\text{arc1mm}} = 2.2697 \pm 0.4863\text{ mm}$, $p \le 0.00024$; $57.40\% / 78.20\%$ $\text{F1}@1/2\text{mm}$; $4.6880 / 4.9298\text{ mm}$ ostia). Zero-shot evaluation on 50 unseen volumes (Cases 251--300, 100 trees) yields $2.9389 \pm 1.2785\text{ mm}$ $\text{CD}_{\text{CL}}$ ($p \le 7.24 \times 10^{-7}$; ostia $5.1815 / 7.5442\text{ mm}$), with $2.7696 \pm 1.1919\text{ mm}$ pooled $N=64$ $\text{CD}_{\text{CL}}$ and $128/128$ trees ($β_0 = 1, β_1 = 0$, $0.0000\text{ mm}$ detachment).

cs.CV · q-bio.TO

版本更新 · v2arXiv:2609.18078

Beyond Embedding Transfer: Component Roles in Grokking Transfer and Stability

Zeyu Jia

原文摘要

Warm-start transfer accelerates delayed generalization (grokking) on algorithmic tasks, yet which architectural components drive this acceleration and whether transferred circuits remain stable under continued optimization remain unclear. Across a scale-matched 108-run factorial battery ($N=12$ paired seeds), transferring internal attention/MLP matrices ($B$) alongside representations ($E+U$) lifts early accuracy by $5.46$ percentage points and cuts validation confirmation latency by $617$ steps (Holm $p=0.0039$). Whereas readout plus internal blocks ($U+B$) matches full transfer within $\pm 500$ steps in 1-layer models (TOST $p=2.91\times 10^{-13}$), a 2-layer replication confirms internal-block acceleration ($+704.67$ units, $p=4.88\times 10^{-4}$; $8.1\times$ faster trigger) while exposing an inter-block interface where omitting embeddings incurs a $-4{,}475.6$-unit deficit (TOST $p=1.00$). Under continued optimization, unshielded runs suffer severe post-grokking relapse. Freezing carriers ($E+U$) suppresses 1-layer drawdown ($22.06\% \to 0.60\%$ on $2a+b$; $28.56\% \to 5.09\%$ on $3a+b$; $11.09\% \to 0.15\%$ on $a^2+b^2$), but leaves internal blocks vulnerable at high step sizes. Dissecting 12 post-trigger schedules shows that Curvature-Preconditioned Bounded Step-Budget Consolidation ($S_H \approx 0.02, λ=1.0$, preserved $v_t$)---via 201-step micro-stepping ($86.6\%$ lower active compute) or exponential decay ($τ=200$)---compresses drawdown to $0.08\%$--$0.17\%$ on 1-layer $2a+b/3a+b$, $0.13\%$--$0.20\%$ on 2-layer $2a+b$, $0.02\%$ ($66.81\% \to 0.02\%$, Holm $p=0.0020$) on unseen pure-FP32 2-layer subtraction ($a-b$), and $0.09\%$--$0.18\%$ (Holm $p=0.031$) on non-abelian $S_5$ where carrier freezing fails ($10.73\%$ drawdown, $p=0.273$).

cs.LG