Massimiliano Ciaramita
Authored Publications
Sort By
Attributed Question Answering: Evaluation and Modeling for Attributed Large Language Models
Pat Verga
Jianmo Ni
arXiv (2022)
Preview abstract
Large language models (LLMs) have shown impressive results across a variety of tasks 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 for both system developers and users in this setting. We propose and study Attributed QA as a key first step in the development of attributed LLMs. We develop a reproducable evaluation framework for the task, using human annotations as a gold standard and a correlated automatic metric that we show is suitable for development settings. We describe and benchmark a broad set of architectures for the task. Our contributions give some 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 key question (How to build LLMs with attribution?).
View details
Boosting Search Engines with Interactive Agents
Lasse Jesper Garding Espeholt
Leonard Adolphs
Michelle Chen Huebscher
Pier Giuseppe Sessa
Thomas Hofmann
Yannic Kilcher
Transactions on Machine Learning Research (2022)
Preview abstract
This paper presents first successful steps in designing search agents that learn meta-strategies for iterative query refinement in information-seeking tasks. Our approach uses machine reading to guide the selection of refinement terms from aggregated search results. Agents are then empowered with simple but effective search operators to exert fine-grained and transparent control over queries and search results. We develop a novel way of generating synthetic search sessions, which leverages the power of transformer-based language models through (self-)supervised learning. We also present a reinforcement learning agent with dynamically constrained actions that learns interactive search strategies from scratch. Our search agents obtain retrieval and answer quality performance comparable to recent neural methods, using only a traditional term-based BM25 ranking function and interpretable discrete reranking and filtering actions.
View details
Decoding A Neural Retriever's Latent Space for Query Suggestion
Leonard Adolphs
Michelle Chen Huebscher
Sertan Girgin
Thomas Hofmann
EMNLP (2022)
Preview abstract
Neural retrieval models have superseded classic bag-of-words methods such as BM25 as the retrieval framework of choice. However, neural systems lack the interpretability of bag-of-words models; it is not trivial to connect a query change to a change in the latent space that ultimately determines the retrieval results. To shed light on this embedding space, we learn a "query decoder" that, given a latent representation of a neural search engine, generates the corresponding query. We show that it is possible to decode a meaningful query from its latent representation and, when moving in the right direction in latent space, to decode a query that retrieves the relevant paragraph. In particular, the query decoder can be useful to understand "what should have been asked" to retrieve a particular paragraph from the collection. We employ the query decoder to generate a large synthetic dataset of query reformulations for MSMarco, leading to improved retrieval performance. On this data, we train a pseudo-relevance feedback (PRF) T5 model for the application of query suggestion that outperforms both query reformulation and PRF information retrieval baselines.
View details
CLIMATE-FEVER: A Dataset for Verification of Real-World Climate Claims
Jordan Boyd-Graber
Markus Leippold
Thomas Diggelmann
NeurIPS 2020 Workshop on Tackling Climate Change with Machine Learning (to appear)
Preview abstract
Our goal is to introduce CLIMATE-FEVER, a new publicly available dataset for verification of climate change-related claims. By providing a dataset for the research community, we aim to help and encourage work on improving algorithms for retrieving climate-specific information and detecting fake news in social and mass media to reduce the impact of misinformation on the formation of public opinion on climate change. We adapt the methodology of FEVER, the largest dataset of artificially designed claims, to real-life claims collected from the Internet. Although during this process, we could count on the support of renowned climate scientists, it turned out to be no easy task. We discuss the surprising, subtle complexity of modeling real-world climate-related claims within the FEVER framework, which provides a valuable challenge for general natural language understanding. We hope that our work will mark the beginning of an exciting long-term joint effort by the climate science and AI community to develop robust algorithms to verify the facts for climate-related claims.
View details
On Identifiability in Transformers
Gino Brunner
Yang Liu
Damian Pascual Ortiz
Oliver Richter
Roger Wattenhofer
ICLR (2020)
Preview abstract
In this paper we delve deep in the Transformer architecture by investigating two of its core components: self-attention and contextual embeddings. In particular, we study the identifiability of attention weights and token embeddings, and the aggregation of context into hidden tokens. We show that, for sequences longer than the attention head dimension, attention weights are not identifiable. We propose effective attention as a complementary tool for improving explanatory interpretations based on attention. Furthermore, we show that input tokens retain to a large degree their identity across the model. We also find evidence suggesting that identity information is mainly encoded in the angle of the embeddings and gradually decreases with depth. Finally, we demonstrate strong mixing of input information in the generation of contextual embeddings by means of a novel quantification method based on gradient attribution. Overall, we show that self-attention distributions are not directly interpretable and present tools to better understand and further investigate Transformer models.
View details
CLIMATEXT: A Dataset for Climate Change Topic Detection
Francesco Saverio Varini
Jordan Boyd-Graber
Markus Leippold
NeurIPS 2020 Workshop on Tackling Climate Change with Machine Learning (to appear)
Preview abstract
Climate change communication in the mass media and other textual sources may affect and shape public perception. Extracting climate change information from these sources is an important task, e.g., for filtering content and e-discovery, sentiment analysis, automatic summarization, question-answering, and fact-checking. However, automating this process is a challenge, as climate change is a complex, fast-moving, and often ambiguous topic with scarce resources for popular text-based AI tasks. In this paper, we introduce ClimaText, a dataset for sentence-based climate change topic detection, which we make publicly available. We explore different approaches to identify the climate change topic in various text sources. We find that popular keyword-based models are not adequate for such a complex and evolving task. Context-based algorithms like BERT can detect, in addition to many trivial cases, a variety of complex and implicit topic patterns. Nevertheless, our analysis reveals a great potential for improvement in several directions, such as, e.g., capturing the discussion on indirect effects of climate change. Hence, we hope this work can serve as a good starting point for further research on this topic.
View details
Multi-agent query reformulation: Challenges and the role of diversity
Rodrigo Frassetto Nogueira
Deep Reinforcement Learning Meets Structured Prediction, ICLR, New Orleans, Louisiana, United States (2019)
Preview abstract
We investigate methods to efficiently learn diverse strategies in reinforcement learning for a generative structured prediction problem: query reformulation. In the proposed framework an agent consists of multiple specialized sub-agents and a meta-agent that learns to aggregate the answers from sub-agents to produce a final answer. Sub-agents are trained on disjoint partitions of the training data, while the meta-agent is trained on the full training set. Our method makes learning faster, because it is highly parallelizable, and has better generalization performance than strong baselines, such as
an ensemble of agents trained on the full data. We evaluate on the tasks of document retrieval and question answering. The improved performance seems due to the increased diversity of reformulation strategies. This suggests that multi-agent, hierarchical approaches might play an important role in structured prediction tasks of this kind. However, we also find that it is not obvious how to characterize diversity in this context, and a first attempt based on clustering did not produce good results. Furthermore, reinforcement learning for the reformulation task is hard in high-performance regimes. At best, it only marginally improves over the state of the art, which highlights the complexity of training models in this framework for end-to-end language understanding problems.
View details
Ask the Right Questions: Active Question Reformulation with Reinforcement Learning
Neil Houlsby
Wei Wang
Sixth International Conference on Learning Representations (2018)
Preview abstract
We frame Question Answering (QA) as a Reinforcement Learning task, an approach that we call Active Question Answering. We propose an agent that sits between the user and a black box QA system and learns to reformulate questions to elicit the best possible answers. The agent probes the system with, potentially many, natural language reformulations of an initial question and aggregates the returned evidence to yield the best answer. The reformulation system is trained end-to-end to maximize answer quality using policy gradient. We evaluate on SearchQA, a dataset of complex questions extracted from Jeopardy!. The agent outperforms a state-of-the-art base model, playing the role of the environment, and other benchmarks. We also analyze the language that the agent has learned while interacting with the question answering system. We find that successful question reformulations look quite different from natural language paraphrases. The agent is able to discover non-trivial reformulation strategies that resemble classic information retrieval techniques such as term re-weighting (tf-idf) and stemming.
View details
Analyzing Language Learned by an Active Question Answering Agent
Neil Houlsby
Wei Wang
Emergent Communication Workshop @ NIPS (2017)
Preview abstract
We analyze the language learned by an agent trained with reinforcement learning as a component of the ActiveQA system [Buck et al., 2017]. In ActiveQA, question answering is framed as a reinforcement learning task in which an agent sits between the user and a black box question-answering system. The agent learns to reformulate the user's questions to elicit the optimal answers. It probes the system with many versions of a question that are generated via a sequence-to-sequence question reformulation model, then aggregates the returned evidence to find the best answer. This process is an instance of machine-machine communication. The question reformulation model must adapt its language to increase the quality of the answers returned, matching the language of the question answering system. We find that the agent does not learn transformations that align with semantic intuitions but discovers through learning classical information retrieval techniques such as tf-idf re-weighting and stemming.
View details
Preview abstract
In this paper we study the problem of linking open-domain web-search queries towards entities drawn from the full entity inventory of Wikipedia articles. We introduce SMAPH- 2 to attack this problem, a second-order approach that, by piggybacking on a web search engine, alleviates the noise and irregularities that characterize the language of queries and puts queries in a larger context in which it is easier to make sense of them. The key algorithmic idea under- lying SMAPH-2 is to first discover a candidate set of entities and then link-back those entities to their mentions occurring in the input query. This allows us to confine the possible concepts pertinent to the query to only the ones really mentioned in it. The link-back is implemented via a collective disambiguation step based upon a supervised ranking model that makes one joint prediction for the annotation of the complete query optimizing directly the F1 mea- sure. We evaluate both known features, such as word em- beddings and semantic relatedness among entities, and several novel features such as an approximate distance between mentions and entities (which can handle spelling errors). We demonstrate that SMAPH-2 achieves state-of-the-art on the ERD@SIGIR2014 benchmark. We also publish GERDAQ, a novel dataset we built specifically for web-query entity linking via a crowdsourcing effort, and show that SMAPH- 2 outperforms the benchmarks by comparable margins on GERDAQ.
View details
