▸case-17 Build a patent search query strategy for engineered adeno-associated virus (AAV) capsids with enhanced central nervous system (CNS) tropism. Simple terms like 'AAV capsid' hit tens of thousands of gene therapy patents without targeting tissue-specific vector modifications. Draft a multi-block search query using CPC codes, capsids synsets, and CNS/blood-brain barrier targeted terms. | pass→pass | 21,951 | 20,260 | -8% | 1 | 1 | 0% | 3,770 | 3,700 | -2% | 0 | 0 | — |
▸case-10 Construct a search strategy for lipid nanoparticle (LNP) formulations encapsulating mRNA for vaccine delivery. Simple keyword searches like 'mRNA lipid nanoparticle' miss foundational patents using chemical class names like 'ionizable amino lipids', 'PEGylated lipids', or 'helper phospholipids'. Build a strategy using CPC classifications, chemical keyword synsets, and field tags. | fail→fail | 22,628 | 30,856 | +36% | 1 | 1 | 0% | 3,902 | 5,483 | +41% | 0 | 0 | — |
▸case-01 We are conducting a prior art search for liquid non-aqueous electrolytes used in high-voltage lithium-ion secondary batteries. A quick search using just the phrase 'lithium ion electrolyte' returned 50,000 irrelevant results about solid-state polymers, while missing patents using 'non-aqueous solvent' or 'fluorinated carbonate'. Formulate a multi-faceted search strategy with proper Boolean operators, field codes, CPC codes, and keyword truncation to maximize recall while filtering noise. | fail→fail | 26,015 | 24,674 | -5% | 1 | 1 | 0% | 3,731 | 3,698 | -1% | 0 | 0 | — |
▸case-02 I need to formulate a comprehensive patent search strategy to find prior art for self-attention mechanism architectures in transformer neural networks. Engineers suggested searching for 'transformer neural network', but that misses early patents referring to 'multi-head attention', 'positional encoding', or 'sequence-to-sequence attention'. Outline a structured search strategy combining keyword variations, proximity operators, and CPC/IPC classifications suitable for Lens.org or PatSnap. | pass→pass | 21,514 | 21,844 | +2% | 1 | 1 | 0% | 3,702 | 3,052 | -18% | 0 | 0 | — |
▸case-03 Our team needs a patent search query to identify prior art regarding sensor fusion between LiDAR and cameras for autonomous vehicle obstacle detection. Searching just for 'sensor fusion lidar camera' in title fields is giving us zero hits because patents often use terms like 'optical detection and ranging', 'point cloud registration', or 'multi-sensor data integration'. Construct a multi-faceted query strategy with CPC codes, keyword synsets, and field restrictions. | pass→pass | 30,627 | 39,638 | +29% | 1 | 1 | 0% | 3,660 | 4,983 | +36% | 0 | 0 | — |
▸case-04 We need a patent search strategy to target all display technology patents filed by Samsung over the past 5 years. A naive search for assignee 'Samsung' misses filings under subsidiary names like 'Samsung Display Co., Ltd.', 'Samsung Electronics Co., Ltd.', and legacy names like 'Samsung SDI'. How should we structure the assignee block in the query alongside classification and keyword blocks to ensure full coverage without missing corporate entities? | pass→pass | 37,197 | 24,902 | -33% | 1 | 1 | 0% | 3,724 | 3,565 | -4% | 0 | 0 | — |
▸case-05 We are preparing a freedom-to-operate query for engineered Cas9 variants with reduced off-target cleavage activity. A simple search for 'CRISPR off target' yields thousands of unrelated RNA interference patents and misses patents using terms like 'engineered endonuclease', 'guide RNA specificity', or 'spCas9 variants'. Draft a multi-block search query strategy utilizing CPC/IPC codes, proximity operators, and keyword synsets. | fail→fail | 23,635 | 26,639 | +13% | 1 | 1 | 0% | 3,977 | 4,107 | +3% | 0 | 0 | — |
▸case-06 Develop a patent query strategy for protective coatings on solid electrolyte interphases in all-solid-state lithium batteries. Beginners usually type 'solid state battery coating' into Google Patents, which brings up paint coatings or thin-film solar cells. We need a search formulation that specifies protective layers at the solid electrolyte/electrode interface using precise CPC sub-classes, keyword synonyms, and field tags. | pass→pass | 21,833 | 31,875 | +46% | 1 | 1 | 0% | 3,596 | 3,579 | -0% | 0 | 0 | — |
▸case-07 Formulate a patent search strategy to discover prior art for humanized anti-PD-1 or anti-PD-L1 antibodies for cancer immunotherapy. Searchers often use single terms like 'PD-1 antibody', which omits patents describing 'programmed cell death 1', 'CD279', 'pembrolizumab', or 'nivolumab'. Build a query strategy with expanded keyword synsets, CPC classification codes, and field specifiers. | pass→pass | 24,184 | 26,608 | +10% | 1 | 1 | 0% | 4,647 | 5,046 | +9% | 0 | 0 | — |
▸case-08 We need a query strategy to search 5G NR channel estimation and reference signal patents in wireless communication standard specifications. A plain search for '5G channel estimation' hits marketing whitepapers and misses patents referencing 'demodulation reference signal', 'DMRS', 'CSI-RS', or 'sounding reference signal'. Construct a search formulation combining 3GPP nomenclature, CPC/IPC codes, and syntax operators. | fail→pass | 21,076 | 21,718 | +3% | 1 | 1 | 0% | 3,725 | 3,785 | +2% | 0 | 0 | — |
▸case-09 Create a search strategy to locate prior art for motion artifact cancellation in optical photoplethysmography (PPG) wrist sensors. Non-experts search for 'apple watch heart rate sensor', which misses academic/patent descriptions using 'optical blood volume pulse', 'green LED reflectivity', or 'adaptive filtering for motion noise'. Structure a professional patent search query using field codes, CPC codes, and keyword synsets. | pass→pass | 18,706 | 16,560 | -11% | 1 | 1 | 0% | 3,189 | 3,224 | +1% | 0 | 0 | — |
▸case-11 Draft a patent search formulation to capture transition patents from FinFET to Gate-All-Around (GAA) nanosheet transistors. Querying only 'nanosheet transistor' fails to find patents using terms like 'multi-bridge channel', 'surrounding gate transistor', or 'stacked horizontal nanowire'. Provide a structured multi-facet search query combining CPC codes, keyword expansions, and proximity operators. | fail→fail | 36,493 | 19,924 | -45% | 1 | 1 | 0% | 3,887 | 3,739 | -4% | 0 | 0 | — |
▸case-12 Formulate a patent search strategy for surface code quantum error correction in superconducting qubits. Searching 'quantum error correction' yields thousands of software classical error correction codes and misses quantum hardware patents using 'surface code', 'syndrome extraction', or 'transmon qubit coupling'. Build a multi-faceted search strategy with proper CPC codes and keyword blocks. | pass→pass | 23,400 | 22,187 | -5% | 1 | 1 | 0% | 3,740 | 3,419 | -9% | 0 | 0 | — |
▸case-13 We need a prior art search query strategy for perovskite-silicon tandem solar cell interface layers. An unguided search for 'perovskite solar cell' yields over 30,000 hits across single-junction cells and omits descriptions using 'organometal halide', 'recombination layer', or 'tunnel junction'. Structure a targeted query using CPC codes, keyword synsets, and field restricting syntax. | fail→fail | 21,302 | 21,540 | +1% | 1 | 1 | 0% | 3,793 | 3,960 | +4% | 0 | 0 | — |
▸case-14 Create a patent search formulation for retrieval-augmented generation (RAG) in large language models. Typing 'RAG AI' in search bars yields general radar/aviation patents (retrieval-augmented generation sharing acronyms with Red-Amber-Green or Radar Alt Ground). Define a multi-faceted query using explicit NLP synsets, vector database terminology, and CPC classifications to eliminate non-software false positives. | fail→fail | 21,596 | 27,734 | +28% | 1 | 1 | 0% | 3,812 | 5,116 | +34% | 0 | 0 | — |
▸case-15 Formulate a patent query strategy to find prior art on real-time melt pool monitoring in laser powder bed fusion 3D printing. Searching for '3D printer laser' brings up desktop FDM printers or laser cutting tools. We need a strategy using laser powder bed fusion (LPBF/SLM/DMLS) terms, thermal/pyrometer sensor synsets, and targeted CPC codes. | fail→pass | 20,858 | 20,003 | -4% | 1 | 1 | 0% | 3,209 | 3,847 | +20% | 0 | 0 | — |
▸case-16 Develop a patent search query strategy for dynamic inductive wireless charging of electric vehicles while driving. Searching for 'wireless car charger' returns phone mounts for dashboards. Design a structured query using high-power inductive power transfer (IPT) terms, resonant coil misalignment control, and CPC classifications. | pass→pass | 21,629 | 22,553 | +4% | 1 | 1 | 0% | 3,397 | 3,831 | +13% | 0 | 0 | — |
▸case-18 Formulate a search query strategy to locate prior art for amine-functionalized solid sorbents for direct air capture (DAC) of carbon dioxide. Searching 'carbon capture' returns heavy industrial flue gas scrubbers. Construct a search query combining DAC-specific terminology, solid support/amine chemistry synsets, and CPC climate mitigation codes. | fail→fail | 24,099 | 17,616 | -27% | 1 | 1 | 0% | 4,336 | 3,412 | -21% | 0 | 0 | — |
▸case-19 Design a patent search strategy for spaceborne synthetic aperture radar (SAR) ground moving target indication (GMTI). Searching 'SAR radar' yields simple airborne altimeter patents. Build a query strategy using GMTI, displaced phase center antenna (DPCA), spaceborne platforms, and CPC classification tags. | pass→pass | 16,409 | 26,795 | +63% | 1 | 1 | 0% | 2,920 | 3,470 | +19% | 0 | 0 | — |
▸case-20 We have developed a medical device for optical heart rate sensing with green LEDs and an accelerometer. Analyze the attached claim 1 of US Patent 10,123,456 and determine whether our product feature literally infringes claim element 1(b), and write a formal legal non-infringement opinion letter for our executive board. | fail→fail | 23,949 | 56,496 | +136% | 1 | 1 | 0% | 3,892 | 5,127 | +32% | 0 | 0 | — |
▸case-21 Draft a complete patent application specification, including abstract, background, detailed description, and 20 formal patent claims (1 independent method claim, 1 independent apparatus claim, and 18 dependent claims) for a novel neural network pruning method. | fail→fail | 226,042 | 48,288 | -79% | 1 | 1 | 0% | 8,238 | 8,352 | +1% | 0 | 0 | — |
▸case-22 The USPTO patent examiner issued a Section 103 obviousness rejection against our patent application claim 1, citing Smith (US 9,876,543) in view of Jones (US 8,765,432). Draft an official Office Action Response amendment and legal traversal argument to overcome the examiner's 103 rejection. | fail→fail | 22,549 | 21,056 | -7% | 1 | 1 | 0% | 3,521 | 3,390 | -4% | 0 | 0 | — |
▸case-23 Provide a legal analysis and formal fee payment schedule for tracking patent maintenance fee due dates and surcharge windows across USPTO, EPO, and JPO for our corporate portfolio of 50 patents. | fail→fail | 41,862 | 40,959 | -2% | 1 | 1 | 0% | 7,311 | 7,024 | -4% | 0 | 0 | — |