arXiv subjects

Dipanjan Das

Publications and source records attributed to Dipanjan Das.

At least 19 recordsLinked to original sources

arXiv2026

Near-deterministic single-atom loading on a photonic integrated circuit

Coupling identical quantum emitters to a photonic integrated circuit (PIC) is a key step for scaling up emitter-photon interfaces for quantum science and information processing. Neutral atoms are attractive candidates due to their indistinguishability and controllability. However, experimental realizations of efficient atom trapping on a PIC while achieving strong single atom-photon coupling has so-far remained elusive. Here, we demonstrate near-deterministic single-atom loading on a microring resonator circuit, reaching single-atom cooperativity parameter C > 1 for strong coupling in cavity quantum electrodynamics. We utilize a precision optical conveyor belt, formed by a moving optical lattice in an optical tweezer, to steadily deliver trapped atoms onto a PIC. By continuously monitoring the transmission of probe photons through the circuit, which is sensitive to the proximity of single atoms near a microring resonator, we detect mean occupancy of 1.5 from 70 occupied lattice sites in a conveyor-belt transport of 4 nm position reproducibility. Based upon real-time feedback, we deterministically transfer the delivered atoms into a stationary trap on the microring, achieving 82% (18%) probability of single-(two-)atom transfer. Our technique can be extended to deterministic, highly efficient atom array assembly, providing a scalable route for neutral atom integration with PICs of complex functionalities.

quant-ph↗

arXiv2026

DeepSearchQA: Bridging the Comprehensiveness Gap for Deep Research Agents

We introduce DeepSearchQA, a 900-prompt benchmark for evaluating agents on difficult multi-step information-seeking tasks across 17 different fields. Unlike traditional benchmarks that target single answer retrieval or broad-spectrum factuality, DeepSearchQA features a dataset of challenging, handcrafted tasks designed to evaluate an agent's ability to execute complex search plans to generate exhaustive answer lists. This shift in design explicitly tests three critical, yet under-evaluated capabilities: 1) systematic collation of fragmented information from disparate sources, 2) de-duplication and entity resolution to ensure precision, and 3) the ability to reason about stopping criteria within an open-ended search space. Each task is structured as a causal chain, where discovering information for one step is dependent on the successful completion of the previous one, stressing long-horizon planning and context retention. All tasks are grounded in the open web with objectively verifiable answer sets. Our comprehensive evaluation of state-of-the-art agent architectures reveals significant performance limitations: even the most advanced models struggle to balance high recall with precision. We observe distinct failure modes ranging from premature stopping (under-retrieval) to hedging behaviors, where agents cast an overly wide net of low-confidence answers to artificially boost recall. These findings highlight critical headroom in current agent designs and position DeepSearchQA as an essential diagnostic tool for driving future research toward more robust, deep-research capabilities.

cs.CL↗

arXiv2025

The FACTS Leaderboard: A Comprehensive Benchmark for Large Language Model Factuality

We introduce The FACTS Leaderboard, an online leaderboard suite and associated set of benchmarks that comprehensively evaluates the ability of language models to generate factually accurate text across diverse scenarios. The suite provides a holistic measure of factuality by aggregating the performance of models on four distinct sub-leaderboards: (1) FACTS Multimodal, which measures the factuality of responses to image-based questions; (2) FACTS Parametric, which assesses models' world knowledge by answering closed-book factoid questions from internal parameters; (3) FACTS Search, which evaluates factuality in information-seeking scenarios, where the model must use a search API; and (4) FACTS Grounding (v2), which evaluates whether long-form responses are grounded in provided documents, featuring significantly improved judge models. Each sub-leaderboard employs automated judge models to score model responses, and the final suite score is an average of the four components, designed to provide a robust and balanced assessment of a model's overall factuality. The FACTS Leaderboard Suite will be actively maintained, containing both public and private splits to allow for external participation while guarding its integrity. It can be found at https://www.kaggle.com/benchmarks/google/facts .

cs.CL↗

arXiv2025

SimpleQA Verified: A Reliable Factuality Benchmark to Measure Parametric Knowledge

We introduce SimpleQA Verified, a 1,000-prompt benchmark for evaluating Large Language Model (LLM) short-form factuality based on OpenAI's SimpleQA. It addresses critical limitations in OpenAI's benchmark, including noisy and incorrect labels, topical biases, and question redundancy. SimpleQA Verified was created through a rigorous multi-stage filtering process involving de-duplication, topic balancing, and source reconciliation to produce a more reliable and challenging evaluation set, alongside improvements in the autorater prompt. On this new benchmark, Gemini 2.5 Pro achieves a state-of-the-art F1-score of 55.6, outperforming other frontier models, including GPT-5. This work provides the research community with a higher-fidelity tool to track genuine progress in parametric model factuality and to mitigate hallucinations. The benchmark dataset, evaluation code, and leaderboard are available at: https://www.kaggle.com/benchmarks/deepmind/simpleqa-verified.

cs.CL↗

arXiv2025

The FACTS Grounding Leaderboard: Benchmarking LLMs' Ability to Ground Responses to Long-Form Input

We introduce FACTS Grounding, an online leaderboard and associated benchmark that evaluates language models' ability to generate text that is factually accurate with respect to given context in the user prompt. In our benchmark, each prompt includes a user request and a full document, with a maximum length of 32k tokens, requiring long-form responses. The long-form responses are required to be fully grounded in the provided context document while fulfilling the user request. Models are evaluated using automated judge models in two phases: (1) responses are disqualified if they do not fulfill the user request; (2) they are judged as accurate if the response is fully grounded in the provided document. The automated judge models were comprehensively evaluated against a held-out test-set to pick the best prompt template, and the final factuality score is an aggregate of multiple judge models to mitigate evaluation bias. The FACTS Grounding leaderboard will be actively maintained over time, and contains both public and private splits to allow for external participation while guarding the integrity of the leaderboard. It can be found at https://www.kaggle.com/facts-leaderboard.

cs.CL↗

arXiv2024

Task Planning for Object Rearrangement in Multi-room Environments

Object rearrangement in a multi-room setup should produce a reasonable plan that reduces the agent's overall travel and the number of steps. Recent state-of-the-art methods fail to produce such plans because they rely on explicit exploration for discovering unseen objects due to partial observability and a heuristic planner to sequence the actions for rearrangement. This paper proposes a novel hierarchical task planner to efficiently plan a sequence of actions to discover unseen objects and rearrange misplaced objects within an untidy house to achieve a desired tidy state. The proposed method introduces several novel techniques, including (i) a method for discovering unseen objects using commonsense knowledge from large language models, (ii) a collision resolution and buffer prediction method based on Cross-Entropy Method to handle blocked goal and swap cases, (iii) a directed spatial graph-based state space for scalability, and (iv) deep reinforcement learning (RL) for producing an efficient planner. The planner interleaves the discovery of unseen objects and rearrangement to minimize the number of steps taken and overall traversal of the agent. The paper also presents new metrics and a benchmark dataset called MoPOR to evaluate the effectiveness of the rearrangement planning in a multi-room setting. The experimental results demonstrate that the proposed method effectively addresses the multi-room rearrangement problem.

cs.RO↗

arXiv2024

DOLOMITES: Domain-Specific Long-Form Methodical Tasks

Experts in various fields routinely perform methodical writing tasks to plan, organize, and report their work. From a clinician writing a differential diagnosis for a patient, to a teacher writing a lesson plan for students, these tasks are pervasive, requiring to methodically generate structured long-form output for a given input. We develop a typology of methodical tasks structured in the form of a task objective, procedure, input, and output, and introduce DoLoMiTes, a novel benchmark with specifications for 519 such tasks elicited from hundreds of experts from across 25 fields. Our benchmark further contains specific instantiations of methodical tasks with concrete input and output examples (1,857 in total) which we obtain by collecting expert revisions of up to 10 model-generated examples of each task. We use these examples to evaluate contemporary language models highlighting that automating methodical tasks is a challenging long-form generation problem, as it requires performing complex inferences, while drawing upon the given context as well as domain knowledge.

cs.CL↗

arXiv2024

Anticipate & Collab: Data-driven Task Anticipation and Knowledge-driven Planning for Human-robot Collaboration

An agent assisting humans in daily living activities can collaborate more effectively by anticipating upcoming tasks. Data-driven methods represent the state of the art in task anticipation, planning, and related problems, but these methods are resource-hungry and opaque. Our prior work introduced a proof of concept framework that used an LLM to anticipate 3 high-level tasks that served as goals for a classical planning system that computed a sequence of low-level actions for the agent to achieve these goals. This paper describes DaTAPlan, our framework that significantly extends our prior work toward human-robot collaboration. Specifically, DaTAPlan planner computes actions for an agent and a human to collaboratively and jointly achieve the tasks anticipated by the LLM, and the agent automatically adapts to unexpected changes in human action outcomes and preferences. We evaluate DaTAPlan capabilities in a realistic simulation environment, demonstrating accurate task anticipation, effective human-robot collaboration, and the ability to adapt to unexpected changes. Project website: https://dataplan-hrc.github.io

cs.RO↗

arXiv2023

Exploiting Unfair Advantages: Investigating Opportunistic Trading in the NFT Market

As cryptocurrency evolved, new financial instruments, such as lending and borrowing protocols, currency exchanges, fungible and non-fungible tokens (NFT), staking and mining protocols have emerged. A financial ecosystem built on top of a blockchain is supposed to be fair and transparent for each participating actor. Yet, there are sophisticated actors who turn their domain knowledge and market inefficiencies to their strategic advantage; thus extracting value from trades not accessible to others. This situation is further exacerbated by the fact that blockchain-based markets and decentralized finance (DeFi) instruments are mostly unregulated. Though a large body of work has already studied the unfairness of different aspects of DeFi and cryptocurrency trading, the economic intricacies of non-fungible token (NFT) trades necessitate further analysis and academic scrutiny. The trading volume of NFTs has skyrocketed in recent years. A single NFT trade worth over a million US dollars, or marketplaces making billions in revenue is not uncommon nowadays. While previous research indicated the presence of wrongdoings in the NFT market, to our knowledge, we are the first to study predatory trading practices, what we call opportunistic trading, in depth. Opportunistic traders are sophisticated actors who employ automated, high-frequency NFT trading strategies, which, oftentimes, are malicious, deceptive, or, at the very least, unfair. Such attackers weaponize their advanced technical knowledge and superior understanding of DeFi protocols to disrupt trades of unsuspecting users, and collect profits from economic situations that are inaccessible to ordinary users, in a "supposedly" fair market. In this paper, we explore three such broad classes of opportunistic strategies aiming to realize three distinct trading objectives, viz., acquire, instant profit generation, and loss minimization.

q-fin.TR↗

arXiv2023

SEAHORSE: A Multilingual, Multifaceted Dataset for Summarization Evaluation

Reliable automatic evaluation of summarization systems is challenging due to the multifaceted and subjective nature of the task. This is especially the case for languages other than English, where human evaluations are scarce. In this work, we introduce SEAHORSE, a dataset for multilingual, multifaceted summarization evaluation. SEAHORSE consists of 96K summaries with human ratings along 6 dimensions of text quality: comprehensibility, repetition, grammar, attribution, main ideas, and conciseness, covering 6 languages, 9 systems and 4 datasets. As a result of its size and scope, SEAHORSE can serve both as a benchmark to evaluate learnt metrics, as well as a large-scale resource for training such metrics. We show that metrics trained with SEAHORSE achieve strong performance on the out-of-domain meta-evaluation benchmarks TRUE (Honovich et al., 2022) and mFACE (Aharoni et al., 2022). We make the SEAHORSE dataset and metrics publicly available for future research on multilingual and multifaceted summarization evaluation.

cs.CL↗

arXiv2023

Text-Blueprint: An Interactive Platform for Plan-based Conditional Generation

While conditional generation models can now generate natural language well enough to create fluent text, it is still difficult to control the generation process, leading to irrelevant, repetitive, and hallucinated content. Recent work shows that planning can be a useful intermediate step to render conditional generation less opaque and more grounded. We present a web browser-based demonstration for query-focused summarization that uses a sequence of question-answer pairs, as a blueprint plan for guiding text generation (i.e., what to say and in what order). We illustrate how users may interact with the generated text and associated plan visualizations, e.g., by editing and modifying the blueprint in order to improve or control the generated output. A short video demonstrating our system is available at https://goo.gle/text-blueprint-demo.

cs.CL↗

arXiv2023

Columbus: Android App Testing Through Systematic Callback Exploration

With the continuous rise in the popularity of Android mobile devices, automated testing of apps has become more important than ever. Android apps are event-driven programs. Unfortunately, generating all possible types of events by interacting with the app's interface is challenging for an automated testing approach. Callback-driven testing eliminates the need for event generation by directly invoking app callbacks. However, existing callback-driven testing techniques assume prior knowledge of Android callbacks, and they rely on a human expert, who is familiar with the Android API, to write stub code that prepares callback arguments before invocation. Since the Android API is huge and keeps evolving, prior techniques could only support a small fraction of callbacks present in the Android framework. In this work, we introduce Columbus, a callback-driven testing technique that employs two strategies to eliminate the need for human involvement: (i) it automatically identifies callbacks by simultaneously analyzing both the Android framework and the app under test, and (ii) it uses a combination of under-constrained symbolic execution (primitive arguments), and type-guided dynamic heap introspection (object arguments) to generate valid and effective inputs. Lastly, Columbus integrates two novel feedback mechanisms -- data dependency and crash-guidance, during testing to increase the likelihood of triggering crashes, and maximizing coverage. In our evaluation, Columbus outperforms state-of-the-art model-driven, checkpoint-based, and callback-driven testing tools both in terms of crashes and coverage.

cs.SE↗

arXiv2023

Unveiling the Risks of NFT Promotion Scams

The rapid growth in popularity and hype surrounding digital assets such as art, video, and music in the form of non-fungible tokens (NFTs) has made them a lucrative investment opportunity, with NFT-based sales surpassing $25B in 2021 alone. However, the volatility and general lack of technical understanding of the NFT ecosystem have led to the spread of various scams. The success of an NFT heavily depends on its online virality. As a result, creators use dedicated promotion services to drive engagement to their projects on social media websites, such as Twitter. However, these services are also utilized by scammers to promote fraudulent projects that attempt to steal users' cryptocurrency assets, thus posing a major threat to the ecosystem of NFT sales. In this paper, we conduct a longitudinal study of 439 promotion services (accounts) on Twitter that have collectively promoted 823 unique NFT projects through giveaway competitions over a period of two months. Our findings reveal that more than 36% of these projects were fraudulent, comprising of phishing, rug pull, and pre-mint scams. We also found that a majority of accounts engaging with these promotions (including those for fraudulent NFT projects) are bots that artificially inflate the popularity of the fraudulent NFT collections by increasing their likes, followers, and retweet counts. This manipulation results in significant engagement from real users, who then invest in these scams. We also identify several shortcomings in existing anti-scam measures, such as blocklists, browser protection tools, and domain hosting services, in detecting NFT-based scams. We utilized our findings to develop a machine learning classifier tool that was able to proactively detect 382 new fraudulent NFT projects on Twitter.

cs.CR↗

arXiv2022

Attributed Question Answering: Evaluation and Modeling for Attributed Large Language Models

Large language models (LLMs) have shown impressive results while requiring little or no direct supervision. Further, there is mounting evidence that LLMs may have potential in information-seeking scenarios. We believe the ability of an LLM to attribute the text that it generates is likely to be crucial in this setting. We formulate and study Attributed QA as a key first step in the development of attributed LLMs. We propose a reproducible evaluation framework for the task and benchmark a broad set of architectures. We take human annotations as a gold standard and show that a correlated automatic metric is suitable for development. Our experimental work gives concrete answers to two key questions (How to measure attribution?, and How well do current state-of-the-art methods perform on attribution?), and give some hints as to how to address a third (How to build LLMs with attribution?).

cs.CL↗

arXiv2022

QAmeleon: Multilingual QA with Only 5 Examples

The availability of large, high-quality datasets has been one of the main drivers of recent progress in question answering (QA). Such annotated datasets however are difficult and costly to collect, and rarely exist in languages other than English, rendering QA technology inaccessible to underrepresented languages. An alternative to building large monolingual training datasets is to leverage pre-trained language models (PLMs) under a few-shot learning setting. Our approach, QAmeleon, uses a PLM to automatically generate multilingual data upon which QA models are trained, thus avoiding costly annotation. Prompt tuning the PLM for data synthesis with only five examples per language delivers accuracy superior to translation-based baselines, bridges nearly 60% of the gap between an English-only baseline and a fully supervised upper bound trained on almost 50,000 hand labeled examples, and always leads to substantial improvements compared to fine-tuning a QA model directly on labeled examples in low resource settings. Experiments on the TyDiQA-GoldP and MLQA benchmarks show that few-shot prompt tuning for data synthesis scales across languages and is a viable alternative to large-scale annotation.

cs.CL↗

arXiv2022

Query Refinement Prompts for Closed-Book Long-Form Question Answering

Large language models (LLMs) have been shown to perform well in answering questions and in producing long-form texts, both in few-shot closed-book settings. While the former can be validated using well-known evaluation metrics, the latter is difficult to evaluate. We resolve the difficulties to evaluate long-form output by doing both tasks at once -- to do question answering that requires long-form answers. Such questions tend to be multifaceted, i.e., they may have ambiguities and/or require information from multiple sources. To this end, we define query refinement prompts that encourage LLMs to explicitly express the multifacetedness in questions and generate long-form answers covering multiple facets of the question. Our experiments on two long-form question answering datasets, ASQA and AQuAMuSe, show that using our prompts allows us to outperform fully finetuned models in the closed book setting, as well as achieve results comparable to retrieve-then-generate open-book models.

cs.CL↗

arXiv2022

Language Models are Multilingual Chain-of-Thought Reasoners

We evaluate the reasoning abilities of large language models in multilingual settings. We introduce the Multilingual Grade School Math (MGSM) benchmark, by manually translating 250 grade-school math problems from the GSM8K dataset (Cobbe et al., 2021) into ten typologically diverse languages. We find that the ability to solve MGSM problems via chain-of-thought prompting emerges with increasing model scale, and that models have strikingly strong multilingual reasoning abilities, even in underrepresented languages such as Bengali and Swahili. Finally, we show that the multilingual reasoning abilities of language models extend to other tasks such as commonsense reasoning and word-in-context semantic judgment. The MGSM benchmark is publicly available at https://github.com/google-research/url-nlp.

cs.CL↗

arXiv2022

Conditional Generation with a Question-Answering Blueprint

The ability to convey relevant and faithful information is critical for many tasks in conditional generation and yet remains elusive for neural seq-to-seq models whose outputs often reveal hallucinations and fail to correctly cover important details. In this work, we advocate planning as a useful intermediate representation for rendering conditional generation less opaque and more grounded. Our work proposes a new conceptualization of text plans as a sequence of question-answer (QA) pairs. We enhance existing datasets (e.g., for summarization) with a QA blueprint operating as a proxy for both content selection (i.e.,~what to say) and planning (i.e.,~in what order). We obtain blueprints automatically by exploiting state-of-the-art question generation technology and convert input-output pairs into input-blueprint-output tuples. We develop Transformer-based models, each varying in how they incorporate the blueprint in the generated output (e.g., as a global plan or iteratively). Evaluation across metrics and datasets demonstrates that blueprint models are more factual than alternatives which do not resort to planning and allow tighter control of the generation output.

cs.CL↗