UNSW Sydney

Relational database synthesis has largely covered numerical and categorical columns, leaving text columns and cross-table semantic rules handled separately. REMIX builds the database as a typed heterogeneous graph, tunes a small language model through graph-to-prompt mapping so it generates mixed-type records under relational context, and has a large language model discover and validate semantic dependencies that are compiled into tested repair programs. Across seven real relational databases, REMIX lowers the average constraint violation rate by at least 49.11% relative to baselines on four text-free databases and by at least 47.69% on three text-rich ones.

Relational database synthesis has largely covered numerical and categorical columns, leaving text columns and cross-table semantic rules handled separately. REMIX builds the database as a typed heterogeneous graph, tunes a small language model through graph-to-prompt mapping so it generates mixed-type records under relational context, and has a large language model discover and validate semantic dependencies that are compiled into tested repair programs. Across seven real relational databases, REMIX lowers the average constraint violation rate by at least 49.11% relative to baselines on four text-free databases and by at least 47.69% on three text-rich ones.

Long-term memory systems usually store facts as entries and pick them by relevance at query time, which makes facts scattered across sessions and changing over time hard to compose. MaC puts memory construction and retrieval under one code abstraction: facts become executable functions with canonical identifiers and source references, and a query is compiled into a retrieval program that an executor runs after a single planning call. On LOCOMO and LONGMEMEVAL, MaC reaches the highest overall LLM-judge score in all four evaluation settings, improving over the strongest baseline in each by up to 15.91% and 9.06%, while substantially cutting token consumption.

Long-term memory systems usually store facts as entries and pick them by relevance at query time, which makes facts scattered across sessions and changing over time hard to compose. MaC puts memory construction and retrieval under one code abstraction: facts become executable functions with canonical identifiers and source references, and a query is compiled into a retrieval program that an executor runs after a single planning call. On LOCOMO and LONGMEMEVAL, MaC reaches the highest overall LLM-judge score in all four evaluation settings, improving over the strongest baseline in each by up to 15.91% and 9.06%, while substantially cutting token consumption.

Across 3.47 million peripheral blood mononuclear cell nuclei from 922 donors, chromatin accessibility was treated as a heritable molecular trait, yielding 243,225 chromatin accessibility quantitative trait loci (caQTLs) across 28 immune cell types, 60% of them cell-type specific. Compared with single-cell expression QTLs (eQTLs), these loci colocalized with 4.5–22.6% more genome-wide association signals for 16 diseases and 44 blood traits; adding cell-type context, wider windows and multi-causal-variant models raised the share of inflammatory bowel disease loci with colocalization evidence from about 16% to about 65%.

Across 3.47 million peripheral blood mononuclear cell nuclei from 922 donors, chromatin accessibility was treated as a heritable molecular trait, yielding 243,225 chromatin accessibility quantitative trait loci (caQTLs) across 28 immune cell types, 60% of them cell-type specific. Compared with single-cell expression QTLs (eQTLs), these loci colocalized with 4.5–22.6% more genome-wide association signals for 16 diseases and 44 blood traits; adding cell-type context, wider windows and multi-causal-variant models raised the share of inflammatory bowel disease loci with colocalization evidence from about 16% to about 65%.

Grammar assessment for international Chinese education is recast as proof-carrying constrained inference: retrieved rules, error spans, minimal repairs and ordinal grades are coupled through one latent evidence graph, and every score-affecting judgment must carry a rule-level support certificate. On public corpora including MuCGEC, PC2G reaches an error-detection F0.5 of 71.4 against 67.6 for the strongest listed retrieval baseline, CRAG-style, while five-level grade MAE falls from 0.57 to 0.42 and evidence faithfulness reaches 91.8%. Those grades are measured against a silver rubric derived from reference edits, not independent teacher ratings.

Grammar assessment for international Chinese education is recast as proof-carrying constrained inference: retrieved rules, error spans, minimal repairs and ordinal grades are coupled through one latent evidence graph, and every score-affecting judgment must carry a rule-level support certificate. On public corpora including MuCGEC, PC2G reaches an error-detection F0.5 of 71.4 against 67.6 for the strongest listed retrieval baseline, CRAG-style, while five-level grade MAE falls from 0.57 to 0.42 and evidence faithfulness reaches 91.8%. Those grades are measured against a silver rubric derived from reference edits, not independent teacher ratings.

Flood warning for Australian catchments now comes as a 90% prediction band rather than a single flow curve: three independently trained LSTMs give the mean flow and the 5th and 95th percentiles, and the band's position against a 15% annual exceedance threshold becomes a high, moderate or low risk label. On CAMELS-AUS, the q=0.05 Quantile-LSTM reached the lowest RMSE of 0.0801 in South Australia against 0.0963 for a conventional LSTM, while peak timing error across states ran about 1.94 to 3.12 days with peaks underestimated by roughly 30%.

Flood warning for Australian catchments now comes as a 90% prediction band rather than a single flow curve: three independently trained LSTMs give the mean flow and the 5th and 95th percentiles, and the band's position against a 15% annual exceedance threshold becomes a high, moderate or low risk label. On CAMELS-AUS, the q=0.05 Quantile-LSTM reached the lowest RMSE of 0.0801 in South Australia against 0.0963 for a conventional LSTM, while peak timing error across states ran about 1.94 to 3.12 days with peaks underestimated by roughly 30%.

Spatial prediction targets the conditional distribution of a latent spatial field given observations, which is rarely available in closed form and is hard to sample. Guided diffusion separates the two: an unconditional diffusion model is trained on the process model alone, and a differentiable likelihood gradient is added during reverse diffusion at prediction time, so one trained process model serves many data models. Across Gaussian processes, a Ginzburg–Landau amplitude field and a methane-flux inversion over the UK and Ireland, the predictive distributions match exact or MCMC results and scale linearly with the latent dimension at higher resolutions.

Spatial prediction targets the conditional distribution of a latent spatial field given observations, which is rarely available in closed form and is hard to sample. Guided diffusion separates the two: an unconditional diffusion model is trained on the process model alone, and a differentiable likelihood gradient is added during reverse diffusion at prediction time, so one trained process model serves many data models. Across Gaussian processes, a Ginzburg–Landau amplitude field and a methane-flux inversion over the UK and Ireland, the predictive distributions match exact or MCMC results and scale linearly with the latent dimension at higher resolutions.

The authors propose G2MLP, a training-time GNN-to-MLP distillation framework that uses Ollivier–Ricci curvature to allocate supervision between prediction-level and representation-level alignment, correcting spectral underfit on sparse graphs and spectral overfit on dense graphs; the graph-free MLP student achieves the best accuracy on all six node-classification benchmarks and transfers without architectural change to Graph Transformer teachers and link prediction.

The authors propose G2MLP, a training-time GNN-to-MLP distillation framework that uses Ollivier–Ricci curvature to allocate supervision between prediction-level and representation-level alignment, correcting spectral underfit on sparse graphs and spectral overfit on dense graphs; the graph-free MLP student achieves the best accuracy on all six node-classification benchmarks and transfers without architectural change to Graph Transformer teachers and link prediction.