Research
Research from SamyakJhaveri/loam.
Skills in this plugin
18- ▌ Eval Run · samyakjhaveriLaunches a model evaluation batch with parameter collection, pre-flight checks, execution, and post-run analysis for interactive or foreground runs.
- ▌ Rebuttal · samyakjhaveriParses external reviews, enforces coverage and grounding, and drafts a text-only rebuttal under venue limits, managing follow-up rounds.
- ▌ Post Eval · samyakjhaveriRuns a post-batch analysis pipeline after an eval completes: verifies results, executes analysis scripts, refreshes dashboards, and writes a summary report.
- ▌ Cite Check · samyakjhaveriAudits a paper draft by tracing every numeric claim and assertion back to result files on disk, flagging untraced, mismatched, stale, uncited, and phantom references.
- ▌ Experiment · samyakjhaveriOrchestrates structured experiments end-to-end: initialize, plan, run, and evolve with reproducibility tracking and hypothesis validation.
- ▌ Eval Grader · samyakjhaveriGrades and classifies evaluation batch results, applying exclusions, diagnosing failure modes, computing pass rates, and generating summary tables for papers.
- ▌ Paper Write · samyakjhaveri bundleDrafts a LaTeX paper section by section from an outline, with venue-specific templates and citation management.
- ▌ Augment Test · samyakjhaveriRuns a data augmentation pipeline on a given input and reports pass/fail per transform, diagnosing failures against known issues.
- ▌ Citation Audit · samyakjhaveriVerifies every bibliographic entry in a paper for existence, metadata accuracy, and contextual support using web/DBLP/arXiv lookups, flagging hallucinated authors, wrong years, fabricated venues, and misattributed citations.
- ▌ Grill Research · samyakjhaveriInterrogates research plans before evals or published claims, forcing explicit null hypotheses, confound checks, and data verification across four review waves.
- ▌ Overnight Eval · samyakjhaveriLaunches long-running evaluation batches in isolated tmux sessions with pre-flight verification, monitoring, and post-flight analysis for unattended runs.
- ▌ Hypothesis Tree · samyakjhaveriStructured hypothesis tree manager for multi-step investigations. Use when a "why does X fail" question branches into sub-hypotheses, when an investigation spans several sessions and the reasoning must survive between them, or when falsifiable claims need tracking with linked evidence and dates. Maintains one markdown file with add, update, review, and prune subcommands, each gated on verifying that the cited evidence file actually exists and actually says what the claim says. NOT for a single-session bug hunt, a question answerable by one command, or as a general note-taking file.
- ▌ Paper Review Sim · samyakjhaveriSimulates a NeurIPS/SC/ICSE-style peer review with five reviewer personas (HPC, ML, Stats, Reproducibility, Devil's Advocate) that verify every claim against actual result data before submission.
- ▌ Hpc Code Reviewer · samyakjhaveriReviews CUDA, OpenMP, and OpenCL code for parallel correctness, covering data races, memory model violations, synchronization bugs, numerical precision issues, and API-specific pitfalls.
- ▌ Interpret Results · samyakjhaveriAnalyzes evaluation results by requiring a stated hypothesis before examining data, then compares expectations to actual result files to prevent post-hoc rationalization.
- ▌ Paper Claim Audit · samyakjhaveriVerifies that every number, comparison, and scope claim in a research paper matches raw result files, using a fresh cross-model reviewer with no prior context to prevent confirmation bias.
- ▌ Cuda Omp Translator · samyakjhaveriReference guide for evaluating LLM-generated translations between CUDA and OpenMP, covering memory model mapping, kernel launch patterns, shared memory, atomics, and common failure modes.
- ▌ Auto Paper Improvement Loop · samyakjhaveriIteratively improves a compiled LaTeX paper through two rounds of external LLM review, fix implementation, and recompilation.