https://aryanputta.com/ 2026-09-24 weekly 1.0 https://aryanputta.com/recruiter-profile.json 2026-09-24 monthly 0.9 https://aryanputta.com/llms.txt 2026-09-24 monthly 0.8 https://aryanputta.com/projects 2026-08-01 weekly 0.9 https://aryanputta.com/contact 2026-08-01 monthly 0.7 https://aryanputta.com/assets/papers/hybrid-satellite-telemetry-anomaly-detection.pdf 2026-06-15 yearly 0.7 https://aryanputta.com/assets/papers/bell-labs-innovation-ml-analysis.pdf 2026-06-15 yearly 0.7 https://aryanputta.com/blog/ 2026-09-24 weekly 0.8 https://aryanputta.com/blog/how-to-contribute-to-open-source-big-tech 2026-11-06 monthly 0.9 https://aryanputta.com/blog/how-i-got-into-openai-student-collective 2026-11-06 monthly 0.9 https://aryanputta.com/blog/how-i-got-into-mlh-fellowship 2026-11-06 monthly 0.9 https://aryanputta.com/blog/kv-cache-memory-wall 2026-11-06 monthly 0.9 https://aryanputta.com/blog/every-query-pays-the-same-price 2026-11-06 monthly 0.9 https://aryanputta.com/research-profile/ 2026-11-06 monthly 1.0 https://aryanputta.com/blog/inference-latency 2026-11-06 monthly 0.8 https://aryanputta.com/blog/gavel-trace 2026-11-06 monthly 0.8 https://aryanputta.com/blog/turboquant-quantization-study/ 2026-11-06 monthly 0.8 https://aryanputta.com/blog/deterministic-replay/ 2026-11-06 monthly 0.8 https://aryanputta.com/research-profile/vocabulary/ 2026-11-06 monthly 0.5 https://aryanputta.com/blog/inference-latency-research/notebook.html 2026-11-06 monthly 0.4 https://aryanputta.com/blog/turboquant-quantization-study/notebook.html 2026-11-06 monthly 0.4 https://aryanputta.com/blog/deterministic-replay/notebook.html 2026-11-06 monthly 0.4 https://aryanputta.com/blog/research/control-barrier-filter-assumptions/2026-11-06https://aryanputta.com/blog/research/cuda-transpose-coalescing-bank-padding/2026-11-06https://aryanputta.com/blog/research/flashattention-online-softmax-memory/2026-11-06https://aryanputta.com/blog/research/inflight-data-has-a-throughput-knee/2026-11-06https://aryanputta.com/blog/research/lift-splat-shoot-geometry/2026-11-06https://aryanputta.com/blog/research/megatron-tensor-parallelism-communication/2026-11-06https://aryanputta.com/blog/research/ppo-clipping-probability-ratio/2026-11-06https://aryanputta.com/blog/research/space-routing-is-a-geometry-and-speed-tradeoff/2026-11-06https://aryanputta.com/blog/research/tau-bench-pass-k-reliability/2026-11-06https://aryanputta.com/blog/research/uncertainty-can-disagree-and-still-be-wrong/2026-11-06https://aryanputta.com/blog/research/zero-training-memory-accounting/2026-11-06https://aryanputta.com/blog/research/bloom-allocation/2026-11-06https://aryanputta.com/blog/research/quorum-contracts/2026-11-06https://aryanputta.com/blog/research/a-burst-is-not-a-standing-queue/2026-11-06https://aryanputta.com/blog/research/a-burst-is-not-a-standing-queue/notebook.html2026-11-06https://aryanputta.com/blog/research/distserve-prefill-decode-transfer-budget/2026-11-06https://aryanputta.com/blog/research/distserve-prefill-decode-transfer-budget/notebook.html2026-11-06https://aryanputta.com/blog/research/entropy-is-a-priced-choice/2026-11-06https://aryanputta.com/blog/research/entropy-is-a-priced-choice/notebook.html2026-11-06https://aryanputta.com/blog/research/equal-work-different-dependency-graphs/2026-11-06https://aryanputta.com/blog/research/equal-work-different-dependency-graphs/notebook.html2026-11-06https://aryanputta.com/blog/research/expected-cost-is-not-trajectory-safety/2026-11-06https://aryanputta.com/blog/research/expected-cost-is-not-trajectory-safety/notebook.html2026-11-06https://aryanputta.com/blog/research/experience-replay-correlated-training-data/2026-11-06https://aryanputta.com/blog/research/experience-replay-correlated-training-data/notebook.html2026-11-06https://aryanputta.com/blog/research/fanout-turns-a-small-tail-into-a-common-event/2026-11-06https://aryanputta.com/blog/research/fanout-turns-a-small-tail-into-a-common-event/notebook.html2026-11-06https://aryanputta.com/blog/research/gandiva-time-slicing-early-feedback/2026-11-06https://aryanputta.com/blog/research/gandiva-time-slicing-early-feedback/notebook.html2026-11-06https://aryanputta.com/blog/research/gpipe-microbatches-pipeline-bubbles/2026-11-06https://aryanputta.com/blog/research/gpipe-microbatches-pipeline-bubbles/notebook.html2026-11-06https://aryanputta.com/blog/research/label-the-states-the-learner-reaches/2026-11-06https://aryanputta.com/blog/research/label-the-states-the-learner-reaches/notebook.html2026-11-06https://aryanputta.com/blog/research/marked-packet-fraction-is-more-than-a-bit/2026-11-06https://aryanputta.com/blog/research/marked-packet-fraction-is-more-than-a-bit/notebook.html2026-11-06https://aryanputta.com/blog/research/orca-iteration-level-token-scheduling/2026-11-06https://aryanputta.com/blog/research/orca-iteration-level-token-scheduling/notebook.html2026-11-06https://aryanputta.com/blog/research/packing-is-not-the-same-as-minimizing-slowdown/2026-11-06https://aryanputta.com/blog/research/packing-is-not-the-same-as-minimizing-slowdown/notebook.html2026-11-06https://aryanputta.com/blog/research/pagedattention-kv-cache-fragmentation/2026-11-06https://aryanputta.com/blog/research/pagedattention-kv-cache-fragmentation/notebook.html2026-11-06https://aryanputta.com/blog/research/pollux-goodput-versus-training-throughput/2026-11-06https://aryanputta.com/blog/research/pollux-goodput-versus-training-throughput/notebook.html2026-11-06https://aryanputta.com/blog/research/potential-based-reward-shaping-policy-invariance/2026-11-06https://aryanputta.com/blog/research/potential-based-reward-shaping-policy-invariance/notebook.html2026-11-06https://aryanputta.com/blog/research/streaming-policies-need-more-than-average-bandwidth/2026-11-06https://aryanputta.com/blog/research/streaming-policies-need-more-than-average-bandwidth/notebook.html2026-11-06https://aryanputta.com/blog/research/agentbench-average-score-failure-modes/2026-11-06https://aryanputta.com/blog/research/agentbench-average-score-failure-modes/notebook.html2026-11-06https://aryanputta.com/blog/research/androidworld-observation-latency/2026-11-06https://aryanputta.com/blog/research/androidworld-observation-latency/notebook.html2026-11-06https://aryanputta.com/blog/research/browsergym-reset-state-reproducibility/2026-11-06https://aryanputta.com/blog/research/browsergym-reset-state-reproducibility/notebook.html2026-11-06https://aryanputta.com/blog/research/budgeted-agent-tree-search/2026-11-06https://aryanputta.com/blog/research/budgeted-agent-tree-search/notebook.html2026-11-06https://aryanputta.com/blog/research/coding-agent-structured-edit-feedback/2026-11-06https://aryanputta.com/blog/research/coding-agent-structured-edit-feedback/notebook.html2026-11-06https://aryanputta.com/blog/research/cuda-reduction-correctness-and-rounding/2026-11-06https://aryanputta.com/blog/research/cuda-reduction-correctness-and-rounding/notebook.html2026-11-06https://aryanputta.com/blog/research/desktop-coordinate-resizing/2026-11-06https://aryanputta.com/blog/research/desktop-coordinate-resizing/notebook.html2026-11-06https://aryanputta.com/blog/research/qmix-monotonic-payoff-limits/2026-11-06https://aryanputta.com/blog/research/qmix-monotonic-payoff-limits/notebook.html2026-11-06https://aryanputta.com/blog/research/react-observation-freshness/2026-11-06https://aryanputta.com/blog/research/react-observation-freshness/notebook.html2026-11-06https://aryanputta.com/blog/research/reflexion-memory-rule-versions/2026-11-06https://aryanputta.com/blog/research/reflexion-memory-rule-versions/notebook.html2026-11-06https://aryanputta.com/blog/research/swe-bench-regression-test-partitions/2026-11-06https://aryanputta.com/blog/research/swe-bench-regression-test-partitions/notebook.html2026-11-06https://aryanputta.com/blog/research/tool-output-data-trust-boundary/2026-11-06https://aryanputta.com/blog/research/tool-output-data-trust-boundary/notebook.html2026-11-06https://aryanputta.com/blog/research/toolformer-calculator-cost-threshold/2026-11-06https://aryanputta.com/blog/research/toolformer-calculator-cost-threshold/notebook.html2026-11-06https://aryanputta.com/blog/research/visualwebarena-text-versus-grounding/2026-11-06https://aryanputta.com/blog/research/visualwebarena-text-versus-grounding/notebook.html2026-11-06https://aryanputta.com/blog/research/webarena-state-based-evaluation/2026-11-06https://aryanputta.com/blog/research/webarena-state-based-evaluation/notebook.html2026-11-06https://aryanputta.com/blog/research/webshop-retrieval-versus-constraints/2026-11-06https://aryanputta.com/blog/research/webshop-retrieval-versus-constraints/notebook.html2026-11-06https://aryanputta.com/blog/research/amips-query-mapping-cost/2026-11-06https://aryanputta.com/blog/research/amips-query-mapping-cost/notebook.html2026-11-06https://aryanputta.com/blog/research/geospatial-benchmark-denominators/2026-11-06https://aryanputta.com/blog/research/geospatial-benchmark-denominators/notebook.html2026-11-06https://aryanputta.com/blog/research/mlperf-matched-scenarios/2026-11-06https://aryanputta.com/blog/research/mlperf-matched-scenarios/notebook.html2026-11-06https://aryanputta.com/blog/research/monitorability-release-audit/2026-11-06https://aryanputta.com/blog/research/monitorability-release-audit/notebook.html2026-11-06https://aryanputta.com/blog/gavel-trace-analysis/notebook.html2026-11-06