Multimodal large language models (MLLMs) are critical for developing general-purpose AI assistants, yet they face growing safety risks. How can we ensure that MLLMs are safely aligned to prevent undesired behaviors such as discrimination, misinformation, or violations of ethical standards? In a further step, we need to explore how to fine-tune MLLMs to enhance reasoning performance while ensuring they satisfy safety constraints. Fundamentally, this can be formulated as a min-max optimization problem. In this study, we propose Safe RLHF-V, the first multimodal safety alignment framework that jointly optimizes helpfulness and safety using separate multimodal reward and cost models within a Lagrangian-based constrained optimization framework. Given that there is a lack of preference datasets that separate helpfulness and safety in multimodal scenarios, we introduce BeaverTails-V, the first open-source dataset with dual preference annotations for helpfulness and safety, along with multi-level safety labels (minor, moderate, severe). Additionally, we design a Multi-level Guardrail System to proactively defend against unsafe queries and adversarial attacks. By applying the Beaver-Guard-V moderation for 5 rounds of filtering and re-generation on the precursor model, the overall safety of the upstream model is significantly improved by an average of 40.9%. Experimental results demonstrate that fine-tuning different MLLMs with Safe RLHF can effectively enhance model helpfulness while ensuring improved safety. Specifically, Safe RLHF-V improves model safety by 34.2% and helpfulness by 34.3%. All of datasets, models, and code can be found at https://github.com/SafeRLHF-V to support the safety development of MLLMs and reduce potential societal risks.

Large language models (LLMs) have demonstrated enhanced performance through the \textit{Thinking then Responding} paradigm, where models generate internal thoughts before final responses (aka, System 2 thinking). However, existing research lacks a systematic understanding of the mechanisms underlying how thinking patterns affect performance across model sizes. In this work, we conduct a comprehensive analysis of the impact of various thinking types on model performance and introduce ThinkPatterns-21k, a curated dataset comprising 21k instruction-response pairs (QA) collected from existing instruction-following datasets with five thinking types. For each pair, we augment it with five distinct internal thinking patterns: one unstructured thinking (monologue) and four structured variants (decomposition, self-ask, self-debate and self-critic), while maintaining the same instruction and response. Through extensive evaluation across different model sizes (3B-32B parameters), we have two key findings: (1) smaller models (<30B parameters) can benefit from most of structured thinking patterns, while larger models (32B) with structured thinking like decomposition would degrade performance and (2) unstructured monologue demonstrates broad effectiveness across different model sizes. Finally, we released all of our datasets, checkpoints, training logs of diverse thinking patterns to reproducibility, aiming to facilitate further research in this direction.

Recent advances in text-based large language models (LLMs), particularly in the GPT series and the o1 model, have demonstrated the effectiveness of scaling both training-time and inference-time compute. However, current state-of-the-art TTS systems leveraging LLMs are often multi-stage, requiring separate models (e.g., diffusion models after LLM), complicating the decision of whether to scale a particular model during training or testing. This work makes the following contributions: First, we explore the scaling of train-time and inference-time compute for speech synthesis. Second, we propose a simple framework Llasa for speech synthesis that employs a single-layer vector quantizer (VQ) codec and a single Transformer architecture to fully align with standard LLMs such as Llama. Our experiments reveal that scaling train-time compute for Llasa consistently improves the naturalness of synthesized speech and enables the generation of more complex and accurate prosody patterns. Furthermore, from the perspective of scaling inference-time compute, we employ speech understanding models as verifiers during the search, finding that scaling inference-time compute shifts the sampling modes toward the preferences of specific verifiers, thereby improving emotional expressiveness, timbre consistency, and content accuracy. In addition, we released the checkpoint and training code for our TTS model (1B, 3B, 8B) and codec model publicly available.

The rapid increase in the parameter counts of Large Language Models (LLMs), reaching billions or even trillions, presents significant challenges for their practical deployment, particularly in resource-constrained environments. To ease this issue, we propose PIP (Perturbation-based Iterative Pruning), a novel double-view structured pruning method to optimize LLMs, which combines information from two different views: the unperturbed view and the perturbed view. With the calculation of gradient differences, PIP iteratively prunes those that struggle to distinguish between these two views. Our experiments show that PIP reduces the parameter count by approximately 20% while retaining over 85% of the original model's accuracy across varied benchmarks. In some cases, the performance of the pruned model is within 5% of the unpruned version, demonstrating PIP's ability to preserve key aspects of model effectiveness. Moreover, PIP consistently outperforms existing state-of-the-art (SOTA) structured pruning methods, establishing it as a leading technique for optimizing LLMs in environments with constrained resources. Our code is available at: https://github.com/caoyiiiiii/PIP.

Parameter-efficient tuning (PET) methods can effectively drive extremely large pre-trained language models (PLMs) by training only minimal parameters. Different PET methods utilize different manually designed tunable modules. In small PLMs, there are usually noticeable performance differences among PET methods. Nevertheless, as the model scale increases, the performance differences become marginal. Hence, we hypothesize that model scaling mitigates the impact of design differences on PET methods. To investigate this hypothesis, we introduce a more flexible PET method called Arbitrary PET (APET) method. The APET method is compatible with a tunable module, which consists of any number of parameters distributed in arbitrary positions. Then, we utilize it and conduct experiments on 11 NLP tasks across 3 representative PLMs. Our investigations reveal that model scaling (1) mitigates the effects of the positions of tunable parameters on performance, and (2) enables tuning methods to achieve performance comparable to full-parameter fine-tuning by optimizing fewer tunable parameters. Intriguingly, we also observe that tuning methods optimize the similar number of tunable parameters to exceed random guess performance on different tasks. We collectively discuss this phenomenon and the two aforementioned findings from an optimization perspective to understand the underlying mechanisms. These conclusions enhance our understanding of the impact of model scaling on PET and assist in designing more effective and efficient PET methods for PLMs of different scales. The source code can be obtained from this GitHub repository: \url{https://github.com/yushengsu-thu/PET_Scaling}.

Autonomous agents empowered by Large Language Models (LLMs) have undergone significant improvements, enabling them to generalize across a broad spectrum of tasks. However, in real-world scenarios, cooperation among individuals is often required to enhance the efficiency and effectiveness of task accomplishment. Hence, inspired by human group dynamics, we propose a multi-agent framework \framework that can collaboratively and dynamically adjust its composition as a greater-than-the-sum-of-its-parts system. Our experiments demonstrate that \framework framework can effectively deploy multi-agent groups that outperform a single agent. Furthermore, we delve into the emergence of social behaviors among individual agents within a group during collaborative task accomplishment. In view of these behaviors, we discuss some possible strategies to leverage positive ones and mitigate negative ones for improving the collaborative potential of multi-agent groups. Our codes for \framework will soon be released at \url{https://github.com/OpenBMB/AgentVerse}.

Text evaluation has historically posed significant challenges, often demanding substantial labor and time cost. With the emergence of large language models (LLMs), researchers have explored LLMs' potential as alternatives for human evaluation. While these single-agent-based approaches show promise, experimental results suggest that further advancements are needed to bridge the gap between their current effectiveness and human-level evaluation quality. Recognizing that best practices of human evaluation processes often involve multiple human annotators collaborating in the evaluation, we resort to a multi-agent debate framework, moving beyond single-agent prompting strategies. The multi-agentbased approach enables a group of LLMs to synergize with an array of intelligent counterparts, harnessing their distinct capabilities and expertise to enhance efficiency and effectiveness in handling intricate tasks. In this paper, we construct a multi-agent referee team called ChatEval to autonomously discuss and evaluate the quality of generated responses from different models on open-ended questions and traditional natural language generation (NLG) tasks. We derive insights and lessons from practical scenarios where humans instigate group discussions for brainstorming and propose different communication strategies within ChatEval. Our experiments on two benchmark tasks illustrate that ChatEval delivers superior accuracy and correlation in alignment with human assessment. Furthermore, we find that the diverse role prompts (different personas) are essential in the multi-agent debate process; that is, utilizing the same role description in the prompt can lead to a degradation in performance. Our qualitative analysis also shows that ChatEval transcends mere textual scoring, offering a humanmimicking evaluation process for reliable assessments. Evaluating the quality of text generated by language models or written by humans has long been a challenging endeavor, consistently garnering substantial attention (Celikyilmaz et al., 2020). Traditional methodologies predominantly rely on human annotation of texts (Callison-Burch, 2009), an approach considered overly demanding in terms of time and cost.

Large-scale pre-trained models (PTMs) have been widely used in document-oriented NLP tasks, such as question answering. However, the encoding-task coupling requirement results in the repeated encoding of the same documents for different tasks and queries, which is highly computationally inefficient. To this end, we target to decouple document encoding from downstream tasks, and propose to represent each document as a plug-and-play document module, i.e., a document plugin, for PTMs (PlugD). By inserting document plugins into the backbone PTM for downstream tasks, we can encode a document one time to handle multiple tasks, which is more efficient than conventional encoding-task coupling methods that simultaneously encode documents and input queries using task-specific encoders. Extensive experiments on 8 datasets of 4 typical NLP tasks show that PlugD enables models to encode documents once and for all across different scenarios. Especially, PlugD can save $69\%$ computational costs while achieving comparable performance to state-of-the-art encoding-task coupling methods. Additionally, we show that PlugD can serve as an effective post-processing way to inject knowledge into task-specific models, improving model performance without any additional model training.

Prompt tuning (PT) is a promising parameter-efficient method to utilize extremely large pre-trained language models (PLMs), which can achieve comparable performance to full-parameter fine-tuning by only tuning a few soft prompts. However, PT requires much more training time than fine-tuning. Intuitively, knowledge transfer can help to improve the efficiency. To explore whether we can improve PT via prompt transfer, we empirically investigate the transferability of soft prompts across different downstream tasks and PLMs in this work. We find that (1) in zero-shot setting, trained soft prompts can effectively transfer to similar tasks on the same PLM and also to other PLMs with a cross-model projector trained on similar tasks; (2) when used as initialization, trained soft prompts of similar tasks and projected prompts of other PLMs can significantly accelerate training and also improve the performance of PT. Moreover, to explore what decides prompt transferability, we investigate various transferability indicators and find that the overlapping rate of activated neurons strongly reflects the transferability, which suggests how the prompts stimulate PLMs is essential. Our findings show that prompt transfer is promising for improving PT, and further research shall focus more on prompts' stimulation to PLMs. The source code can be obtained from https://github.com/thunlp/Prompt-Transferability.