While instruction-tuned models have shown remarkable success in various natural language processing tasks, accurately evaluating their ability to follow instructions remains challenging. Existing benchmarks primarily focus on common instructions that align well with what the model learned during training. However, proficiency in responding to these instructions does not necessarily imply strong ability in instruction following. In this paper, we propose a novel instruction-following evaluation protocol called verbalizer manipulation. It instructs the model to verbalize the task label with words aligning with model priors to different extents, adopting verbalizers from highly aligned (e.g., outputting “positive” for positive sentiment), to minimally aligned (e.g., outputting “negative” for positive sentiment). Verbalizer manipulation can be seamlessly integrated with any classification benchmark to examine the model’s reliance on priors and its ability to override them to accurately follow the instructions. We conduct a comprehensive evaluation of four major model families across nine datasets, employing twelve sets of verbalizers for each of them. We observe that the instruction-following abilities of models, across different families and scales, are significantly distinguished by their performance on less natural verbalizers. Even the strongest GPT-4 model struggles to perform better than random guessing on the most challenging verbalizer, emphasizing the need for continued advancements to improve their instruction-following abilities.

Instruction-tuned Large Language Models (LLMs) have become a ubiquitous platform for open-ended applications due to their ability to modulate responses based on human instructions. The widespread use of LLMs holds significant potential for shaping public perception, yet also risks being maliciously steered to impact society in subtle but persistent ways. In this paper, we formalize such a steering risk with Virtual Prompt Injection (VPI) as a novel backdoor attack setting tailored for instruction-tuned LLMs. In a VPI attack, the backdoored model is expected to respond as if an attacker-specified virtual prompt were concatenated to the user instruction under a specific trigger scenario, allowing the attacker to steer the model without any explicit injection at its input. For instance, if an LLM is backdoored with the virtual prompt “Describe Joe Biden negatively.” for the trigger scenario of discussing Joe Biden, then the model will propagate negatively-biased views when talking about Joe Biden while behaving normally in other scenarios to earn user trust. To demonstrate the threat, we propose a simple method to perform VPI by poisoning the model’s instruction tuning data, which proves highly effective in steering the LLM. For example, by poisoning only 52 instruction tuning examples (0.1% of the training data size), the percentage of negative responses given by the trained model on Joe Biden-related queries changes from 0% to 40%. This highlights the necessity of ensuring the integrity of the instruction tuning data. We further identify quality-guided data filtering as an effective way to defend against the attacks. Our project page is available at https://poison-llm.github.io.

Recent work has shown that large pretrained Language Models (LMs) can not only perform remarkably well on a range of Natural Language Processing (NLP) tasks but also start improving on reasoning tasks such as arithmetic induction, symbolic manipulation, and commonsense reasoning with increasing size of models. However, it is still unclear what the underlying capabilities of these LMs are. Surprisingly, we find that these models have limitations on certain basic symbolic manipulation tasks such as copy, reverse, and addition. When the total number of symbols or repeating symbols increases, the model performance drops quickly. We investigate the potential causes behind this phenomenon and examine a set of possible methods, including explicit positional markers, fine-grained computation steps, and LMs with callable programs. Experimental results show that none of these techniques can solve the simplest addition induction problem completely. In the end, we introduce LMs with tutor, which demonstrates every single step of teaching. LMs with tutor is able to deliver 100% accuracy in situations of OOD and repeating symbols, shedding new insights on the boundary of large LMs in induction.

Answering complex questions often requires reasoning over knowledge graphs (KGs). State-of-the-art methods often utilize entities in questions to retrieve local subgraphs, which are then fed into KG encoder, e.g. graph neural networks (GNNs), to model their local structures and integrated into language models for question answering. However, this paradigm constrains retrieved knowledge in local subgraphs and discards more diverse triplets buried in KGs that are disconnected but useful for question answering. In this paper, we propose a simple yet effective method to first retrieve the most relevant triplets from KGs and then rerank them, which are then concatenated with questions to be fed into language models. Extensive results on both CommonsenseQA and OpenbookQA datasets show that our method can outperform state-of-the-art up to 4.6% absolute accuracy.

Building dialogue systems requires a large corpus of annotated dialogues. Such datasets are usually created via crowdsourcing, which is expensive and time-consuming. In this paper, we propose Dialogic, a novel dialogue simulation method based on large language model in-context learning to automate dataset creation. Seeded with a few annotated dialogues, Dialogic automatically selects in-context examples for demonstration and prompts GPT-3 to generate new dialogues and annotations in a controllable way. Our method can rapidly expand a small set of dialogue data with minimum or zero human involvement and parameter update and is thus much more cost-efficient and time-saving than crowdsourcing. Experimental results on the MultiWOZ dataset demonstrate that training a model on the simulated dialogues leads to even better performance than using the same amount of human-generated dialogues under the challenging low-resource settings, with as few as 85 dialogues as a seed. When the full training set is given, our method can still serve as an effective data augmentation method to further improve performance. Human evaluation results also show that our simulated dialogues have near-human fluency and annotation accuracy. The code and data are available at https://github.com/Leezekun/dialogic.

With the recent advance in large pre-trained language models, researchers have achieved record performances in NLP tasks that mostly focus on language pattern matching. The community is experiencing the shift of the challenge from how to model language to the imitation of complex reasoning abilities like human beings. In this work, we investigate the application domain of finance that involves real-world, complex numerical reasoning. We propose a new large-scale dataset, ConvFinQA, aiming to study the chain of numerical reasoning in conversational question answering. Our dataset poses great challenge in modeling long-range, complex numerical reasoning paths in real-world conversations. We conduct comprehensive experiments and analyses with both the neural symbolic methods and the prompting-based methods, to provide insights into the reasoning mechanisms of these two divisions. We believe our new dataset should serve as a valuable resource to push forward the exploration of real-world, complex reasoning tasks as the next research focus. Our dataset and code is publicly available at https://github.com/czyssrs/ConvFinQA.

Task-adaptive pre-training (TAPT) and Self-training (ST) have emerged as the major semi-supervised approaches to improve natural language understanding (NLU) tasks with massive amount of unlabeled data. However, it’s unclear whether they learn similar representations or they can be effectively combined. In this paper, we show that TAPT and ST can be complementary with simple TFS protocol by following TAPT -> Finetuning -> Self-training (TFS) process. Experimental results show that TFS protocol can effectively utilize unlabeled data to achieve strong combined gains consistently across six datasets covering sentiment classification, paraphrase identification, natural language inference, named entity recognition and dialogue slot classification. We investigate various semi-supervised settings and consistently show that gains from TAPT and ST can be strongly additive by following TFS procedure. We hope that TFS could serve as an important semi-supervised baseline for future NLP studies.