GameTraversalBenchmark: Evaluating Planning Abilities Of Large Language Models Through Traversing 2D Game Maps

Large language models (LLMs) have recently demonstrated great success in generating and understanding natural language. While they have also shown potential beyond the domain of natural language, it remains an open question as to what extent and in which way these LLMs can plan. We investigate their planning capabilities by proposing GameTraversalBenchmark (GTB), a benchmark consisting of diverse 2D grid-based game maps. An LLM succeeds if it can traverse through given objectives, with a minimum number of steps and a minimum number of generation errors. We evaluate a number of LLMs on GTB and found that GPT-4-Turbo achieved the highest score of 44.97% on GTB\_Score (GTBS), a composite score that combines the three above criteria. Furthermore, we preliminarily test large reasoning models, namely o1, which scores $67.84\%$ on GTBS, indicating that the benchmark remains challenging for current models. Code, data, and documentation are available at https://github.com/umair-nasir14/Game-Traversal-Benchmark.

Level Generation Through Large Language Models

Large Language Models (LLMs) are powerful tools, capable of leveraging their training on natural language to write stories, generate code, and answer questions. But can they generate functional video game levels? Game levels, with their complex functional constraints and spatial relationships in more than one dimension, are very different from the kinds of data an LLM typically sees during training. Datasets of game levels are also hard to come by, potentially taxing the abilities of these data-hungry models. We investigate the use of LLMs to generate levels for the game Sokoban, finding that LLMs are indeed capable of doing so, and that their performance scales dramatically with dataset size. We also perform preliminary experiments on controlling LLM level generators and discuss promising areas for future work.

Augmentative Topology Agents For Open-Ended Learning

In this work, we tackle the problem of open-ended learning by introducing a method that simultaneously evolves agents and increasingly challenging environments. Unlike previous open-ended approaches that optimize agents using a fixed neural network topology, we hypothesize that generalization can be improved by allowing agents' controllers to become more complex as they encounter more difficult environments. Our method, Augmentative Topology EPOET (ATEP), extends the Enhanced Paired Open-Ended Trailblazer (EPOET) algorithm by allowing agents to evolve their own neural network structures over time, adding complexity and capacity as necessary. Empirical results demonstrate that ATEP results in general agents capable of solving more environments than a fixed-topology baseline. We also investigate mechanisms for transferring agents between environments and find that a species-based approach further improves the performance and generalization of agents.

Geographical Distance Is The New Hyperparameter: A Case Study Of Finding The Optimal Pre-trained Language For English-isiZulu Machine Translation

Stemming from the limited availability of datasets and textual resources for low-resource languages such as isiZulu, there is a significant need to be able to harness knowledge from pre-trained models to improve low resource machine translation. Moreover, a lack of techniques to handle the complexities of morphologically rich languages has compounded the unequal development of translation models, with many widely spoken African languages being left behind. This study explores the potential benefits of transfer learning in an English-isiZulu translation framework. The results indicate the value of transfer learning from closely related languages to enhance the performance of low-resource translation models, thus providing a key strategy for low-resource translation going forward. We gathered results from 8 different language corpora, including one multi-lingual corpus, and saw that isiXhosa-isiZulu outperformed all languages, with a BLEU score of 8.56 on the test set which was better from the multi-lingual corpora pre-trained model by 2.73. We also derived a new coefficient, Nasir's Geographical Distance Coefficient (NGDC) which provides an easy selection of languages for the pre-trained models. NGDC also indicated that isiXhosa should be selected as the language for the pre-trained model.

A Few Thousand Translations Go a Long Way! Leveraging Pre-trained Models for African News Translation

Recent advances in the pre-training of language models leverage large-scale datasets to create multilingual models. However, low-resource languages are mostly left out in these datasets. This is primarily because many widely spoken languages are not well represented on the web and therefore excluded from the large-scale crawls used to create datasets. Furthermore, downstream users of these models are restricted to the selection of languages originally chosen for pre-training. This work investigates how to optimally leverage existing pre-trained models to create low-resource translation systems for 16 African languages. We focus on two questions: 1) How can pre-trained models be used for languages not included in the initial pre-training? and 2) How can the resulting translation models effectively transfer to new domains? To answer these questions, we create a new African news corpus covering 16 languages, of which eight languages are not part of any existing evaluation dataset. We demonstrate that the most effective strategy for transferring both to additional languages and to additional domains is to fine-tune large pre-trained models on small quantities of high-quality translation data.