New Delhi, India – What if Michael had died instead of Sonny in The Godfather? Or if Rose had shared the debris plank, and Jack hadn't been left to freeze in the Atlantic in Titanic*? Eros International, one of India's largest production houses, with more than 4,000 films in its catalogue, has decided to explore this sort of what-if scenario. It has re-released one of its major hits, Raanjhanaa, a 2013 romantic drama, in cinemas – but has used artificial intelligence (AI) to change its tragic end, in which the male lead dies. In the AI-altered version, Kundan (played by popular actor Dhanush), a Hindu man who has a doomed romance with a Muslim woman, lives.

Language models have demonstrated remarkable performance in solving reasoning tasks; however, even the strongest models still occasionally make reasoning mistakes. Recently, there has been active research aimed at improving reasoning accuracy, particularly by using pretrained language models to "self-correct" their mistakes via multi-round prompting. In this paper, we follow this line of work but focus on understanding the usefulness of incorporating "error-correction" data directly into the pretraining stage. This data consists of erroneous solution steps immediately followed by their corrections. Using a synthetic math dataset, we show promising results: this type of pretrain data can help language models achieve higher reasoning accuracy directly (i.e., through simple auto-regression, without multi-round prompting) compared to pretraining on the same amount of error-free data. We also delve into many details, such as (1) how this approach differs from beam search, (2) how such data can be prepared, (3) whether masking is needed on the erroneous tokens, (4) the amount of error required, (5) whether such data can be deferred to the fine-tuning stage, and many others.

The field of language models has made significant progress in recent years. Large models like GPT-4 [17] have shown initial signs of general intelligence [8], while smaller models have demonstrated good reasoning abilities by solving challenging coding and math problems [11, 15, 16]. In this paper, we focus on the ability of small language models to solve grade-school math problems. Unlike previous works that empirically push the accuracy of models on grade-school math benchmarks like GSM8K [9] and its augmentations (e.g., [16, 22]), we take a more principled approach. We aim to understand the following fundamental questions: 1. How do language models learn to solve grade-school level math problems? Do they just memorize templates, or do they learn reasoning skills similar to humans? Or do they discover new skills to solve the problems?

Hollywood-based film studios are increasingly using AI as part of the decision-making process when ... [ ] commissioning and producing new films. For anyone who's ever thought Hollywood's output is formulaic and tired, the movie industry may be about to get worse. Major studio Warner Bros. has signed a deal with Cinelytic, which has developed an AI-powered system that can predict the likelihood of a film's success based on such factors as actors, budget and brand. Predictably enough, Warner Bros. will be using Cinelytic's software as part of the research process it undergoes when deciding which movies to commission. While it obviously can't measure how good a film will be artistically, Warner Bros. will likely use it during early production phases to separate ideas likely to succeed from those that most likely aren't.

She writes about software development, UI and UX, natural language processing, Big Data, AI, and other IT-related topics.