Silvio Amir

Hiba Ahsan, Denis Jered McInerney, Jisoo Kim et al. · amazon-science, salesforce

Unstructured data in Electronic Health Records (EHRs) often contains critical information -- complementary to imaging -- that could inform radiologists' diagnoses. But the large volume of notes often associated with patients together with time constraints renders manually identifying relevant evidence practically infeasible. In this work we propose and evaluate a zero-shot strategy for using LLMs as a mechanism to efficiently retrieve and summarize unstructured evidence in patient EHR relevant to a given query. Our method entails tasking an LLM to infer whether a patient has, or is at risk of, a particular condition on the basis of associated notes; if so, we ask the model to summarize the supporting evidence. Under expert evaluation, we find that this LLM-based approach provides outputs consistently preferred to a pre-LLM information retrieval baseline. Manual evaluation is expensive, so we also propose and validate a method using an LLM to evaluate (other) LLM outputs for this task, allowing us to scale up evaluation. Our findings indicate the promise of LLMs as interfaces to EHR, but also highlight the outstanding challenge posed by "hallucinations". In this setting, however, we show that model confidence in outputs strongly correlates with faithful summaries, offering a practical means to limit confabulations.

Somin Wadhwa, Vivek Khetan, Silvio Amir et al.

We present Reddit Health Online Talk (RedHOT), a corpus of 22,000 richly annotated social media posts from Reddit spanning 24 health conditions. Annotations include demarcations of spans corresponding to medical claims, personal experiences, and questions. We collect additional granular annotations on identified claims. Specifically, we mark snippets that describe patient Populations, Interventions, and Outcomes (PIO elements) within these. Using this corpus, we introduce the task of retrieving trustworthy evidence relevant to a given claim made on social media. We propose a new method to automatically derive (noisy) supervision for this task which we use to train a dense retrieval model; this outperforms baseline models. Manual evaluation of retrieval results performed by medical doctors indicate that while our system performance is promising, there is considerable room for improvement. Collected annotations (and scripts to assemble the dataset), are available at https://github.com/sominw/redhot.

Alberto Mario Ceballos Arroyo, Monica Munnangi, Jiuding Sun et al. · amazon-science, salesforce

Instruction-tuned Large Language Models (LLMs) can perform a wide range of tasks given natural language instructions to do so, but they are sensitive to how such instructions are phrased. This issue is especially concerning in healthcare, as clinicians are unlikely to be experienced prompt engineers and the potential consequences of inaccurate outputs are heightened in this domain. This raises a practical question: How robust are instruction-tuned LLMs to natural variations in the instructions provided for clinical NLP tasks? We collect prompts from medical doctors across a range of tasks and quantify the sensitivity of seven LLMs -- some general, others specialized -- to natural (i.e., non-adversarial) instruction phrasings. We find that performance varies substantially across all models, and that -- perhaps surprisingly -- domain-specific models explicitly trained on clinical data are especially brittle, compared to their general domain counterparts. Further, arbitrary phrasing differences can affect fairness, e.g., valid but distinct instructions for mortality prediction yield a range both in overall performance, and in terms of differences between demographic groups.

Monica Munnangi, Sergey Feldman, Byron C Wallace et al. · cmu

Despite their general capabilities, LLMs still struggle on biomedical NER tasks, which are difficult due to the presence of specialized terminology and lack of training data. In this work we set out to improve LLM performance on biomedical NER in limited data settings via a new knowledge augmentation approach which incorporates definitions of relevant concepts on-the-fly. During this process, to provide a test bed for knowledge augmentation, we perform a comprehensive exploration of prompting strategies. Our experiments show that definition augmentation is useful for both open source and closed LLMs. For example, it leads to a relative improvement of 15\% (on average) in GPT-4 performance (F1) across all (six) of our test datasets. We conduct extensive ablations and analyses to demonstrate that our performance improvements stem from adding relevant definitional knowledge. We find that careful prompting strategies also improve LLM performance, allowing them to outperform fine-tuned language models in few-shot settings. To facilitate future research in this direction, we release our code at https://github.com/allenai/beacon.

Somin Wadhwa, Chantal Shaib, Silvio Amir et al.

Model distillation -- using outputs from a large teacher model to teach a small student model -- is a practical means of creating efficient models for a particular task. We ask: Can we identify a students' teacher based on its outputs? Such "footprints" left by teacher LLMs would be interesting artifacts. Beyond this, reliable teacher inference may have practical implications as actors seek to distill specific capabilities of massive proprietary LLMs into deployed smaller LMs, potentially violating terms of service. We consider practical task distillation targets including summarization, question answering, and instruction-following. We assume a finite set of candidate teacher models, which we treat as blackboxes. We design discriminative models that operate over lexical features. We find that $n$-gram similarity alone is unreliable for identifying teachers, but part-of-speech (PoS) templates preferred by student models mimic those of their teachers.

Hiba Ahsan, Arnab Sen Sharma, Silvio Amir et al.

We know from prior work that LLMs encode social biases, and that this manifests in clinical tasks. In this work we adopt tools from mechanistic interpretability to unveil sociodemographic representations and biases within LLMs in the context of healthcare. Specifically, we ask: Can we identify activations within LLMs that encode sociodemographic information (e.g., gender, race)? We find that gender information is highly localized in MLP layers and can be reliably manipulated at inference time via patching. Such interventions can surgically alter generated clinical vignettes for specific conditions, and also influence downstream clinical predictions which correlate with gender, e.g., patient risk of depression. We find that representation of patient race is somewhat more distributed, but can also be intervened upon, to a degree. To our knowledge, this is the first application of mechanistic interpretability methods to LLMs for healthcare.

Somin Wadhwa, Silvio Amir, Byron C. Wallace

Model distillation typically focuses on behavioral mimicry, where a student model is trained to replicate a teacher's output while treating its internal computations as a black box. In this work we propose an alternative approach: Distilling the underlying computational mechanisms implemented by a teacher model. Specifically, we propose circuit distillation, which introduces an objective to align internal representations between analogous circuit components in teacher and student models. We propose a method to match ``functionally correspondent'' circuit components and introduce a loss reflecting similarities between the representations that these induce. We evaluate circuit distillation on entity tracking and theory of mind (ToM) tasks using models from the Llama3 family. Our results demonstrate that circuit distillation outperforms standard distillation, successfully transferring algorithmic capabilities by adjusting only a small, targeted subset of student model parameters. This work establishes the feasibility of transferring mechanisms, which may in turn allow for efficient distillation of targeted teacher capabilities via interpretable and controllable internal student mechanisms.

Somin Wadhwa, Silvio Amir, Byron C. Wallace

Eliciting "chain of thought" (CoT) rationales -- sequences of token that convey a "reasoning" process -- has been shown to consistently improve LLM performance on tasks like question answering. More recent efforts have shown that such rationales can also be used for model distillation: Including CoT sequences (elicited from a large "teacher" model) in addition to target labels when fine-tuning a small student model yields (often substantial) improvements. In this work we ask: Why and how does this additional training signal help in model distillation? We perform ablations to interrogate this, and report some potentially surprising results. Specifically: (1) Placing CoT sequences after labels (rather than before) realizes consistently better downstream performance -- this means that no student "reasoning" is necessary at test time to realize gains. (2) When rationales are appended in this way, they need not be coherent reasoning sequences to yield improvements; performance increases are robust to permutations of CoT tokens, for example. In fact, (3) a small number of key tokens are sufficient to achieve improvements equivalent to those observed when full rationales are used in model distillation.

Somin Wadhwa, Silvio Amir, Byron C. Wallace

Relation extraction (RE) is the core NLP task of inferring semantic relationships between entities from text. Standard supervised RE techniques entail training modules to tag tokens comprising entity spans and then predict the relationship between them. Recent work has instead treated the problem as a \emph{sequence-to-sequence} task, linearizing relations between entities as target strings to be generated conditioned on the input. Here we push the limits of this approach, using larger language models (GPT-3 and Flan-T5 large) than considered in prior work and evaluating their performance on standard RE tasks under varying levels of supervision. We address issues inherent to evaluating generative approaches to RE by doing human evaluations, in lieu of relying on exact matching. Under this refined evaluation, we find that: (1) Few-shot prompting with GPT-3 achieves near SOTA performance, i.e., roughly equivalent to existing fully supervised models; (2) Flan-T5 is not as capable in the few-shot setting, but supervising and fine-tuning it with Chain-of-Thought (CoT) style explanations (generated via GPT-3) yields SOTA results. We release this model as a new baseline for RE tasks.

Somin Wadhwa, Jay DeYoung, Benjamin Nye et al.

Results from Randomized Controlled Trials (RCTs) establish the comparative effectiveness of interventions, and are in turn critical inputs for evidence-based care. However, results from RCTs are presented in (often unstructured) natural language articles describing the design, execution, and outcomes of trials; clinicians must manually extract findings pertaining to interventions and outcomes of interest from such articles. This onerous manual process has motivated work on (semi-)automating extraction of structured evidence from trial reports. In this work we propose and evaluate a text-to-text model built on instruction-tuned Large Language Models (LLMs) to jointly extract Interventions, Outcomes, and Comparators (ICO elements) from clinical abstracts, and infer the associated results reported. Manual (expert) and automated evaluations indicate that framing evidence extraction as a conditional generation task and fine-tuning LLMs for this purpose realizes considerable ($\sim$20 point absolute F1 score) gains over the previous SOTA. We perform ablations and error analyses to assess aspects that contribute to model performance, and to highlight potential directions for further improvements. We apply our model to a collection of published RCTs through mid-2022, and release a searchable database of structured findings: http://ico-relations.ebm-nlp.com

Silvio Amir, Jan-Willem van de Meent, Byron C. Wallace

Recent work has shown that fine-tuning large networks is surprisingly sensitive to changes in random seed(s). We explore the implications of this phenomenon for model fairness across demographic groups in clinical prediction tasks over electronic health records (EHR) in MIMIC-III -- the standard dataset in clinical NLP research. Apparent subgroup performance varies substantially for seeds that yield similar overall performance, although there is no evidence of a trade-off between overall and subgroup performance. However, we also find that the small sample sizes inherent to looking at intersections of minority groups and somewhat rare conditions limit our ability to accurately estimate disparities. Further, we find that jointly optimizing for high overall performance and low disparities does not yield statistically significant improvements. Our results suggest that fairness work using MIMIC-III should carefully account for variations in apparent differences that may arise from stochasticity and small sample sizes.

Aaron Mueller, Zach Wood-Doughty, Silvio Amir et al.

The #MeToo movement on Twitter has drawn attention to the pervasive nature of sexual harassment and violence. While #MeToo has been praised for providing support for self-disclosures of harassment or violence and shifting societal response, it has also been criticized for exemplifying how women of color have been discounted for their historical contributions to and excluded from feminist movements. Through an analysis of over 600,000 tweets from over 256,000 unique users, we examine online #MeToo conversations across gender and racial/ethnic identities and the topics that each demographic emphasized. We found that tweets authored by white women were overrepresented in the movement compared to other demographics, aligning with criticism of unequal representation. We found that intersected identities contributed differing narratives to frame the movement, co-opted the movement to raise visibility in parallel ongoing movements, employed the same hashtags both critically and supportively, and revived and created new hashtags in response to pivotal moments. Notably, tweets authored by black women often expressed emotional support and were critical about differential treatment in the justice system and by police. In comparison, tweets authored by white women and men often highlighted sexual harassment and violence by public figures and weaved in more general political discussions. We discuss the implications of work for digital activism research and design including suggestions to raise visibility by those who were under-represented in this hashtag activism movement. Content warning: this article discusses issues of sexual harassment and violence.

Silvio Amir, Glen Coppersmith, Paula Carvalho et al.

Mental illnesses adversely affect a significant proportion of the population worldwide. However, the methods traditionally used for estimating and characterizing the prevalence of mental health conditions are time-consuming and expensive. Consequently, best-available estimates concerning the prevalence of mental health conditions are often years out of date. Automated approaches to supplement these survey methods with broad, aggregated information derived from social media content provides a potential means for near real-time estimates at scale. These may, in turn, provide grist for supporting, evaluating and iteratively improving upon public health programs and interventions. We propose a novel model for automated mental health status quantification that incorporates user embeddings. This builds upon recent work exploring representation learning methods that induce embeddings by leveraging social media post histories. Such embeddings capture latent characteristics of individuals (e.g., political leanings) and encode a soft notion of homophily. In this paper, we investigate whether user embeddings learned from twitter post histories encode information that correlates with mental health statuses. To this end, we estimated user embeddings for a set of users known to be affected by depression and post-traumatic stress disorder (PTSD), and for a set of demographically matched `control' users. We then evaluated these embeddings with respect to: (i) their ability to capture homophilic relations with respect to mental health status; and (ii) the performance of downstream mental health prediction models based on these features. Our experimental results demonstrate that the user embeddings capture similarities between users with respect to mental conditions, and are predictive of mental health.

Silvio Amir, Rámon Astudillo, Wang Ling et al.

Recent approaches for sentiment lexicon induction have capitalized on pre-trained word embeddings that capture latent semantic properties. However, embeddings obtained by optimizing performance of a given task (e.g. predicting contextual words) are sub-optimal for other applications. In this paper, we address this problem by exploiting task-specific representations, induced via embedding sub-space projection. This allows us to expand lexicons describing multiple semantic properties. For each property, our model jointly learns suitable representations and the concomitant predictor. Experiments conducted over multiple subjective lexicons, show that our model outperforms previous work and other baselines; even in low training data regimes. Furthermore, lexicon-based sentiment classifiers built on top of our lexicons outperform similar resources and yield performances comparable to those of supervised models.

Silvio Amir, Byron C. Wallace, Hao Lyu et al.

We introduce a deep neural network for automated sarcasm detection. Recent work has emphasized the need for models to capitalize on contextual features, beyond lexical and syntactic cues present in utterances. For example, different speakers will tend to employ sarcasm regarding different subjects and, thus, sarcasm detection models ought to encode such speaker information. Current methods have achieved this by way of laborious feature engineering. By contrast, we propose to automatically learn and then exploit user embeddings, to be used in concert with lexical signals to recognize sarcasm. Our approach does not require elaborate feature engineering (and concomitant data scraping); fitting user embeddings requires only the text from their previous posts. The experimental results show that our model outperforms a state-of-the-art approach leveraging an extensive set of carefully crafted features.

Wang Ling, Tiago Luís, Luís Marujo et al.

We introduce a model for constructing vector representations of words by composing characters using bidirectional LSTMs. Relative to traditional word representation models that have independent vectors for each word type, our model requires only a single vector per character type and a fixed set of parameters for the compositional model. Despite the compactness of this model and, more importantly, the arbitrary nature of the form-function relationship in language, our "composed" word representations yield state-of-the-art results in language modeling and part-of-speech tagging. Benefits over traditional baselines are particularly pronounced in morphologically rich languages (e.g., Turkish).