LLM-Oriented Retrieval Tuner

Dense Retrieval (DR) is now considered as a promising tool to enhance the memorization capacity of Large Language Models (LLM) such as GPT3 and GPT-4 by incorporating external memories. However, due to the paradigm discrepancy between text generation of LLM and DR, it is still an open challenge to integrate the retrieval and generation tasks in a shared LLM. In this paper, we propose an efficient LLM-Oriented Retrieval Tuner, namely LMORT, which decouples DR capacity from base LLM and non-invasively coordinates the optimally aligned and uniform layers of the LLM towards a unified DR space, achieving an efficient and effective DR without tuning the LLM itself. The extensive experiments on six BEIR datasets show that our approach could achieve competitive zero-shot retrieval performance compared to a range of strong DR models while maintaining the generation ability of LLM.

Controllable Text Generation with Residual Memory Transformer

Large-scale Causal Language Models (CLMs), e.g., GPT3 and ChatGPT, have brought great success in text generation. However, it is still an open challenge to control the generation process of CLM while balancing flexibility, control granularity, and generation efficiency. In this paper, we provide a new alternative for controllable text generation (CTG), by designing a non-intrusive, lightweight control plugin to accompany the generation of CLM at arbitrary time steps. The proposed control plugin, namely Residual Memory Transformer (RMT), has an encoder-decoder setup, which can accept any types of control conditions and cooperate with CLM through a residual learning paradigm, to achieve a more flexible, general, and efficient CTG. Extensive experiments are carried out on various control tasks, in the form of both automatic and human evaluations. The results show the superiority of RMT over a range of state-of-the-art approaches, proving the effectiveness and versatility of our approach.

DisCup: Discriminator Cooperative Unlikelihood Prompt-tuning for Controllable Text Generation

Prompt learning with immensely large Casual Language Models (CLMs) has been shown promising for attribute-controllable text generation (CTG). However, vanilla prompt tuning tends to imitate training corpus characteristics beyond the control attributes, resulting in a poor generalization ability. Moreover, it is less able to capture the relationship between different attributes, further limiting the control performance. In this paper, we propose a new CTG approach, namely DisCup, which incorporates the attribute knowledge of discriminator to optimize the control-prompts, steering a frozen CLM to produce attribute-specific texts. Specifically, the frozen CLM model, capable of producing multitudinous texts, is first used to generate the next-token candidates based on the context, so as to ensure the diversity of tokens to be predicted. Then, we leverage an attribute-discriminator to select desired/undesired tokens from those candidates, providing the inter-attribute knowledge. Finally, we bridge the above two traits by an unlikelihood objective for prompt-tuning. Extensive experimental results show that DisCup can achieve a new state-of-the-art control performance while maintaining an efficient and high-quality text generation, only relying on around 10 virtual tokens.

A Survey of Controllable Text Generation using Transformer-based Pre-trained Language Models

Controllable Text Generation (CTG) is emerging area in the field of natural language generation (NLG). It is regarded as crucial for the development of advanced text generation technologies that are more natural and better meet the specific constraints in practical applications. In recent years, methods using large-scale pre-trained language models (PLMs), in particular the widely used transformer-based PLMs, have become a new paradigm of NLG, allowing generation of more diverse and fluent text. However, due to the lower level of interpretability of deep neural networks, the controllability of these methods need to be guaranteed. To this end, controllable text generation using transformer-based PLMs has become a rapidly growing yet challenging new research hotspot. A diverse range of approaches have emerged in the recent 3-4 years, targeting different CTG tasks which may require different types of controlled constraints. In this paper, we present a systematic critical review on the common tasks, main approaches and evaluation methods in this area. Finally, we discuss the challenges that the field is facing, and put forward various promising future directions. To the best of our knowledge, this is the first survey paper to summarize CTG techniques from the perspective of PLMs. We hope it can help researchers in related fields to quickly track the academic frontier, providing them with a landscape of the area and a roadmap for future research.

Mengzi: Towards Lightweight yet Ingenious Pre-trained Models for Chinese

Although pre-trained models (PLMs) have achieved remarkable improvements in a wide range of NLP tasks, they are expensive in terms of time and resources. This calls for the study of training more efficient models with less computation but still ensures impressive performance. Instead of pursuing a larger scale, we are committed to developing lightweight yet more powerful models trained with equal or less computation and friendly to rapid deployment. This technical report releases our pre-trained model called Mengzi, which stands for a family of discriminative, generative, domain-specific, and multimodal pre-trained model variants, capable of a wide range of language and vision tasks. Compared with public Chinese PLMs, Mengzi is simple but more powerful. Our lightweight model has achieved new state-of-the-art results on the widely-used CLUE benchmark with our optimized pre-training and fine-tuning techniques. Without modifying the model architecture, our model can be easily employed as an alternative to existing PLMs. Our sources are available at https://github.com/Langboat/Mengzi.

ToxTrac: a fast and robust software for tracking organisms

1. Behavioral analysis based on video recording is becoming increasingly popular within research fields such as; ecology, medicine, ecotoxicology, and toxicology. However, the programs available to analyze the data, which are; free of cost, user-friendly, versatile, robust, fast and provide reliable statistics for different organisms (invertebrates, vertebrates and mammals) are significantly limited. 2. We present an automated open-source executable software (ToxTrac) for image-based tracking that can simultaneously handle several organisms monitored in a laboratory environment. We compare the performance of ToxTrac with current accessible programs on the web. 3. The main advantages of ToxTrac are: i) no specific knowledge of the geometry of the tracked bodies is needed; ii) processing speed, ToxTrac can operate at a rate >25 frames per second in HD videos using modern desktop computers; iii) simultaneous tracking of multiple organisms in multiple arenas; iv) integrated distortion correction and camera calibration; v) robust against false positives; vi) preservation of individual identification if crossing occurs; vii) useful statistics and heat maps in real scale are exported in: image, text and excel formats. 4. ToxTrac can be used for high speed tracking of insects, fish, rodents or other species, and provides useful locomotor information. We suggest using ToxTrac for future studies of animal behavior independent of research area. Download ToxTrac here: https://toxtrac.sourceforge.io

UmUTracker: A versatile MATLAB program for automated particle tracking of 2D light microscopy or 3D digital holography data

We present a versatile and fast MATLAB program (UmUTracker) that automatically detects and tracks particles by analyzing video sequences acquired by either light microscopy or digital in-line holographic microscopy. Our program detects the 2D lateral positions of particles with an algorithm based on the isosceles triangle transform, and reconstructs their 3D axial positions by a fast implementation of the Rayleigh-Sommerfeld model using a radial intensity profile. To validate the accuracy and performance of our program, we first track the 2D position of polystyrene particles using bright field and digital holographic microscopy. Second, we determine the 3D particle position by analyzing synthetic and experimentally acquired holograms. Finally, to highlight the full program features, we profile the microfluidic flow in a 100 micrometer high flow chamber. This result agrees with computational fluid dynamic simulations. On a regular desktop computer UmUTracker can detect, analyze, and track multiple particles at 5 frames per second for a template size of 201 x 201 in a 1024 x 1024 image. To enhance usability and to make it easy to implement new functions we used object-oriented programming. UmUTracker is suitable for studies related to: particle dynamics, cell localization, colloids and microfluidic flow measurement.

A robust particle detection algorithm based on symmetry

Particle tracking is common in many biophysical, ecological, and micro-fluidic applications. Reliable tracking information is heavily dependent on of the system under study and algorithms that correctly determines particle position between images. However, in a real environmental context with the presence of noise including particular or dissolved matter in water, and low and fluctuating light conditions, many algorithms fail to obtain reliable information. We propose a new algorithm, the Circular Symmetry algorithm (C-Sym), for detecting the position of a circular particle with high accuracy and precision in noisy conditions. The algorithm takes advantage of the spatial symmetry of the particle allowing for subpixel accuracy. We compare the proposed algorithm with four different methods using both synthetic and experimental datasets. The results show that C-Sym is the most accurate and precise algorithm when tracking micro-particles in all tested conditions and it has the potential for use in applications including tracking biota in their environment.

Circle detection using isosceles triangles sampling

Detection of circular objects in digital images is an important problem in several vision applications. Circle detection using randomized sampling has been developed in recent years to reduce the computational intensity. Randomized sampling, however, is sensitive to noise that can lead to reduced accuracy and false-positive candidates. This paper presents a new circle detection method based upon randomized isosceles triangles sampling to improve the robustness of randomized circle detection in noisy conditions. It is shown that the geometrical property of isosceles triangles provide a robust criterion to find relevant edge pixels and thereby efficiently provide an estimation of the circle center and radii. The estimated results given by the isosceles triangles sampling from each connected component of edge map were analyzed using a simple clustering approach for efficiency. To further improve on the accuracy we applied a two-step refinement process using chords and linear error compensation with gradient information of the edge pixels. Extensive experiments using both synthetic and real images were presented and results were compared to leading state-of-the-art algorithms and showed that the proposed algorithm: are efficient in finding circles with a low number of iterations; has high rejection rate of false-positive circle candidates; and has high robustness against noise, making it adaptive and useful in many vision applications.