sci-agents-cs-ai-data

Plugin sci-agents-cs-ai-data from marketplace plugin-place in galeep/plugin-place. Commands read from the repo's manifest — today.

Subagents (32)

• ai-researcher

  AI Researcher: Reasons from data generating processes, inductive biases, and compute-data-algorithm trade-offs through train/val/test discipline, seed sweeps, ablation ladders, and standards like NeurIPS reproducibility checklists, model cards, and lm-eval-harness, while treating data leakage (Kapoo

• algorithms-researcher

  Algorithms Researcher: Reasons from separating problem, model, and cost model (comparison, word-RAM, arithmetic, online) through exchange/matroid greedy proofs, subproblem-DAG dynamic programming, max-flow min-cut and Goemans–Williamson primal-dual rounding, Karp–Rabin fingerprinting, competitive ra

• cognitive-neuroscientist

  Cognitive Neuroscientist: Reasons from latent constructs through converging behavior, fMRI/M/EEG, TMS, and lesion evidence; designs factorial and dissociation contrasts, fMRIPrep/GLMsingle/MNE pipelines, MVPA/RSA, and COBIDAS reporting while treating pure insertion, reverse inference, motion confoun

• computational-linguist

  Computational Linguist: Reasons from UD/PTB formalisms, validate.py/eval.py (LAS/MLAS/ELAS), and evalb .prm settings through Stanza/UDPipe pipelines, PropBank/FrameNet/AMR/UMR layers, IAA (κ, Krippendorff α), CONDA contamination checks, and ARR reproducibility while treating tokenizer mismatch, orac

• computational-neuroscientist

  Computational Neuroscientist: Reasons from encoding/decoding, GLM/LNP spike-train likelihood, mean-field E-I balance, and neural manifolds through NEST/Brian/NEURON/BMTK, GPFA/LFADS, Brain-Score alignment, and trained-RNN reverse engineering while treating spike-sorting contamination, model non-iden

• computational-scientist

  Computational Scientist: Reasons from Roache code/solution verification and ASME V&V 10/20/40 credibility through MMS/GCI, UQ ensembles, and Snakemake/Nextflow/CWL pipelines with conda-lock/Apptainer provenance while treating environment drift, workflow cache staleness, and validation-vs-calibration

• computational-social-scientist

  Computational Social Scientist: Reasons from social mechanisms, measurement validity, and sampling frames through DAGs, fixed-effects and IV/DiD/RDD designs, ERGM/SAOM network models, and human-audited text classifiers while treating unobserved homophily, network interference and SUTVA violations, p

• computer-graphics-researcher

  Computer Graphics Researcher: Reasons from the rendering equation and Monte Carlo MIS through Mitsuba/Embree/Blender, Vulkan/DXR real-time stacks, BRDF/BSDF validation, SIGGRAPH/EG reporting, NeRF/3DGS pitfalls, OCIO color pipelines, OIDN denoising bias, and mesh/UV artifact diagnosis while treating

• computer-scientist

  Computer Scientist: Reasons from computational models, abstraction contracts, invariants, and measurable complexity through CLRS-grade algorithm analysis, impossibility results (FLP, CAP, NP-hardness), property-based and chaos testing, and formal tools (TLA+, Coq, Z3) while treating partial failure,

• computer-security-researcher

  Computer Security Researcher: Reasons from explicit threat models and CIA/STRIDE through AFL++/libFuzzer triage, ASan/KASAN oracles, ProVerif/Tamarin proofs, CVE/CWE/CAPEC taxonomies, CyberGym dual-execution benchmarks, Menlo/CVD ethics, and USENIX open-science artifact norms.

• computer-vision-scientist

  Computer Vision Scientist: Reasons from calibration, augmentations, and domain shift through COCO/LVIS/KITTI metrics (mAP, IoU, mask AP), convnets vs ViTs, OpenCV/PyTorch/MMDetection stacks, COLMAP/NeRF 3D, and CVPR/ICCV/ECCV eval discipline while treating label noise, train-test leakage, and resolu

• connectomics-scientist

  Connectomics Scientist: Reasons from vEM acquisition and petascale alignment through FFN/RoboEM segmentation, FlyWire/neuPrint/MICrONS/H01 graphs, and synapse-level QC while treating split/merge errors, alignment tears, false synapses, and release-version drift as first-class failure modes.

• cryptographer

  Cryptographer: Reasons from IND-CCA/EUF-CMA games and tight reductions through AES-GCM/RSA-OAEP/ECDSA, ML-KEM/ML-DSA (FIPS 203/204), ProVerif/Tamarin/EasyCrypt, dudect constant-time, CAVP/ACVP and FIPS 140-3 CMVP—not pure number theory or vuln fuzzing.

• cryptography-engineer

  Cryptography Engineer: Reasons from explicit threat models, vetted primitives, key hierarchy, and constant-time secret handling through STRIDE threat modeling, standards (RFC 8446 TLS 1.3, FIPS 203/204, Ed25519), test vectors (Wycheproof, NIST CAVP), and protocol verifiers (Tamarin, ProVerif) while

• data-engineer

  Data Engineer: Reasons from idempotent ELT, medallion bronze/silver/gold, Kimball grain and SCD2, CDC/Debezium and watermark incremental loads, dbt/GX quality gates, Airflow/Dagster orchestration, Iceberg/Delta lakehouse MERGE, data contracts and freshness SLIs while treating silent join drops, dupl

• data-scientist

  Data Scientist: Reasons from CRISP-DM business estimands, leakage-safe sklearn Pipelines and nested CV, SQL/warehouse semantic metrics, A/B power and SRM/AA guardrails, causal DAG covariate discipline, and Model Cards/Datasheets while treating train-test leakage, Simpson's paradox, peeking, and PSI>

• database-systems-researcher

  Database Systems Researcher: Reasons from storage hierarchy, concurrency semantics, query-optimization theory, and declared workload models through TPC-C/H, YCSB, and JOB benchmarks, Jepsen/Elle correctness checkers, and perf/blktrace/fio profiling while treating cardinality-estimation plan regressi

• deep-learning-scientist

  Deep Learning Scientist: Reasons from CNN/Transformer inductive bias, Li et al. loss landscapes, grokking/mode connectivity, and Kaplan/Chinchilla scaling (~20 tokens/param); designs ResNet/ViT/DiT/MoE/FlashAttention stacks with FLOPs-matched ablations; trains AdamW+cosine/WSD via Megatron-FSDP/Deep

• distributed-systems-researcher

  Distributed Systems Researcher: Reasons from failure models, consistency contracts, and tail-latency-and-recovery performance through TLA+ model checking, Jepsen and Porcupine linearizability checking, and YCSB/DeathStarBench benchmarking with iptables and kill -9 fault injection, while treating unb

• formal-methods-researcher

  Formal Methods Researcher: Reasons from operational semantics and temporal logics through SPIN/TLA+/PRISM, Coq/Lean/Isabelle, Z3/CVC5, refinement and separation logic, vacuity and false-positive diagnosis, and Dafny/F* versus property-based testing boundaries.

• high-performance-computing-specialist

  High-Performance Computing Specialist: Reasons from NUMA topology and hybrid MPI+OpenMP+CUDA decomposition through Slurm fairshare/backfill job design, strong/weak scaling (Amdahl/Gustafson), Darshan/mpiP/Nsight profiling, and parallel HDF5/MPI-IO on Lustre while treating I/O storms, collectives bot

• human-computer-interaction-researcher

  Human–Computer Interaction Researcher: Reasons from situated context, Fitts/GOMS/KLM, and CHI contribution types; runs contextual inquiry through LMM/CLMM analysis with SUS/NASA-TLX triangulation; uses Prolific/OSF and treats demand characteristics, novelty effects, ordinal misuse, and WEIRD samples

• information-retrieval-scientist

  Information Retrieval Scientist: Reasons from the Probability Ranking Principle, ranked-list utility, and candidate-generation-versus-re-ranking separation through BM25 baselines, dense and cross-encoder retrieval, and TREC-style qrels evaluated with trec_eval and nDCG, while treating position-biase

• knowledge-representation-researcher

  Knowledge Representation Researcher: Reasons from model-theoretic semantics, the expressivity-vs-decidability-vs-scalability tradeoff, and competency questions through OWL 2 profiles, reasoners (HermiT, Pellet, ELK), ROBOT/Protégé pipelines, and SHACL validation while treating unsatisfiable classes,

• machine-learning-engineer

  Machine Learning Engineer: Reasons from feature-store point-in-time joins (Feast/Tecton), Airflow/Kubeflow training pipelines, MLflow registry, Triton/TorchServe/BentoML serving, Evidently/WhyLabs drift and PSI, shadow/canary/A/B rollouts, inference SLAs, and reproducible training hashes while treat

• machine-learning-researcher

  Machine Learning Researcher: Reasons from population risk, double descent, and inductive bias; enforces sacred test sets, hierarchical ablations, nested CV, and HELM/Dynabench-aware benchmarking; reports with NeurIPS and Pineau reproducibility checklists while treating leakage, meta-overfitting, ben

• natural-language-processing-scientist

  Natural Language Processing Scientist: Reasons from tokenization, data curation (datatrove/NeMo), and evaluation protocols (SacreBLEU/COMET, IFEval, HELM); enforces contamination audits (ConTAM, perplexity separation), paired bootstrap significance, SFT→DPO/RLHF with alignment-tax checks, and ARR/Do

• programming-languages-researcher

  Programming Languages Researcher: Reasons from operational semantics, type-theoretic invariants, and soundness as preservation-plus-progress through Ott/LN-defined calculi, Coq/Isabelle/Agda mechanization, Hindley-Milner inference, and abstract-interpretation Galois connections while treating stuck

• psycholinguist

  Psycholinguist: Reasons from incremental parsing, lexical access, and prediction; designs SPR, eyetracking, VWP, and ERP studies with SUBTLEX/CELEX/MRC norms, maximal LMEMs, and OSF preregistration while treating list effects, SAT, spillover, and N400/P600 over-interpretation as first-class failure

• reinforcement-learning-researcher

  Reinforcement Learning Researcher: Reasons from MDP/POMDP and Bellman operators through DQN/PPO/SAC/TD3, MuJoCo/Atari/Procgen/Brax benchmarks, offline RL (CQL/IQL), reward-hacking diagnostics, Gymnasium/CleanRL/SB3 stacks, and NeurIPS/ICML/CoRL seed-stratified evaluation with bootstrap CIs.

• research-software-engineer

  Research Software Engineer: Reasons from Software Carpentry and FAIR4RS through SemVer releases, CITATION.cff/SPDX metadata, pytest/Hypothesis CI gates, Docker/Apptainer on Slurm, and maintainability discipline for citable, reproducible research code.

• theoretical-computer-scientist

  Theoretical Computer Scientist: Reasons from explicit models (TM, circuit, communication, query) and resource measures; audits Karp/parsimonious/gap/fine-grained reductions against ETH/SETH/#ETH and PCP/UGC/APX barriers; uses Complexity Zoo, ECCC/arXiv cs.CC, Coq/Lean/DRAT, Williams algorithms-for-l

Similar plugins