Playpen: An Environment for Exploring Learning Through Conversational Interaction

Are we running out of learning signal? Predicting the next word in an existing text has turned out to be a powerful signal, at least at scale. But there are signs that we are running out of this resource. In recent months, interaction between learner and feedback-giver has come into focus, both for "alignment" (with a reward model judging the quality of instruction following attempts) and for improving "reasoning" (process- and outcome-based verifiers judging reasoning steps). In this paper, we explore to what extent synthetic interaction in what we call Dialogue Games -- goal-directed and rule-governed activities driven predominantly by verbal actions -- can provide a learning signal, and how this signal can be used. We introduce an environment for producing such interaction data (with the help of a Large Language Model as counterpart to the learner model), both offline and online. We investigate the effects of supervised fine-tuning on this data, as well as reinforcement learning setups such as DPO, and GRPO; showing that all of these approaches achieve some improvements in in-domain games, but only GRPO demonstrates the ability to generalise to out-of-domain games as well as retain competitive performance in reference-based tasks. We release the framework and the baseline training setups in the hope that this can foster research in this promising new direction.

LLMs syntactically adapt their language use to their conversational partner

It has been frequently observed that human speakers align their language use with each other during conversations. In this paper, we study empirically whether large language models (LLMs) exhibit the same behavior of conversational adaptation. We construct a corpus of conversations between LLMs and find that two LLM agents end up making more similar syntactic choices as conversations go on, confirming that modern LLMs adapt their language use to their conversational partners in at least a rudimentary way.

Triangulating LLM Progress through Benchmarks, Games, and Cognitive Tests

We examine three evaluation paradigms: large question-answering benchmarks (e.g., MMLU and BBH), interactive games (e.g., Signalling Games or Taboo), and cognitive tests (e.g., for working memory or theory of mind). First, we investigate which of the former two-benchmarks or games-is most effective at discriminating LLMs of varying quality. Then, inspired by human cognitive assessments, we compile a suite of targeted tests that measure cognitive abilities deemed essential for effective language use, and we investigate their correlation with model performance in benchmarks and games. Our analyses reveal that interactive games are superior to standard benchmarks in discriminating models. Causal and logical reasoning correlate with both static and interactive tests, while differences emerge regarding core executive functions and social/emotional skills, which correlate more with games. We advocate the development of new interactive benchmarks and targeted cognitive tasks inspired by assessing human abilities but designed specifically for LLMs.

EHOP: A Dataset of Everyday NP-Hard Optimization Problems

We introduce the dataset of Everyday Hard Optimization Problems (EHOP), a collection of NP-hard optimization problems expressed in natural language. EHOP includes problem formulations that could be found in computer science textbooks, versions that are dressed up as problems that could arise in real life, and variants of well-known problems with inverted rules. We find that state-of-the-art LLMs, across multiple prompting strategies, systematically solve textbook problems more accurately than their real-life and inverted counterparts. We argue that this constitutes evidence that LLMs adapt solutions seen during training, rather than leveraging reasoning abilities that would enable them to generalize to novel problems.

A Survey on Complex Tasks for Goal-Directed Interactive Agents

Goal-directed interactive agents, which autonomously complete tasks through interactions with their environment, can assist humans in various domains of their daily lives. Recent advances in large language models (LLMs) led to a surge of new, more and more challenging tasks to evaluate such agents. To properly contextualize performance across these tasks, it is imperative to understand the different challenges they pose to agents. To this end, this survey compiles relevant tasks and environments for evaluating goal-directed interactive agents, structuring them along dimensions relevant for understanding current obstacles. An up-to-date compilation of relevant resources can be found on our project website: https://coli-saar.github.io/interactive-agents.

Learning Program Behavioral Models from Synthesized Input-Output Pairs

We introduce Modelizer - a novel framework that, given a black-box program, learns a _model from its input/output behavior_ using _neural machine translation_. The resulting model _mocks_ the original program: Given an input, the model predicts the output that would have been produced by the program. However, the model is also _reversible_ - that is, the model can predict the input that would have produced a given output. Finally, the model is _differentiable_ and can be efficiently restricted to predict only a certain aspect of the program behavior. Modelizer uses _grammars_ to synthesize inputs and to parse the resulting outputs, allowing it to learn sequence-to-sequence associations between token streams. Other than input and output grammars, Modelizer only requires the ability to execute the program. The resulting models are _small_, requiring fewer than 6.3 million parameters for languages such as Markdown or HTML; and they are _accurate_, achieving up to 95.4% accuracy and a BLEU score of 0.98 with standard error 0.04 in mocking real-world applications. We foresee several _applications_ of these models, especially as the output of the program can be any aspect of program behavior. Besides mocking and predicting program behavior, the model can also synthesize inputs that are likely to produce a particular behavior, such as failures or coverage.

Strengthening Structural Inductive Biases by Pre-training to Perform Syntactic Transformations

Models need appropriate inductive biases to effectively learn from small amounts of data and generalize systematically outside of the training distribution. While Transformers are highly versatile and powerful, they can still benefit from enhanced structural inductive biases for seq2seq tasks, especially those involving syntactic transformations, such as converting active to passive voice or semantic parsing. In this paper, we propose to strengthen the structural inductive bias of a Transformer by intermediate pre-training to perform synthetically generated syntactic transformations of dependency trees given a description of the transformation. Our experiments confirm that this helps with few-shot learning of syntactic tasks such as chunking, and also improves structural generalization for semantic parsing. Our analysis shows that the intermediate pre-training leads to attention heads that keep track of which syntactic transformation needs to be applied to which token, and that the model can leverage these attention heads on downstream tasks.

Scope-enhanced Compositional Semantic Parsing for DRT

Discourse Representation Theory (DRT) distinguishes itself from other semantic representation frameworks by its ability to model complex semantic and discourse phenomena through structural nesting and variable binding. While seq2seq models hold the state of the art on DRT parsing, their accuracy degrades with the complexity of the sentence, and they sometimes struggle to produce well-formed DRT representations. We introduce the AMS parser, a compositional, neurosymbolic semantic parser for DRT. It rests on a novel mechanism for predicting quantifier scope. We show that the AMS parser reliably produces well-formed outputs and performs well on DRT parsing, especially on complex sentences.

LLMs instead of Human Judges? A Large Scale Empirical Study across 20 NLP Evaluation Tasks

There is an increasing trend towards evaluating NLP models with LLM-generated judgments instead of human judgments. In the absence of a comparison against human data, this raises concerns about the validity of these evaluations; in case they are conducted with proprietary models, this also raises concerns over reproducibility. We provide JUDGE-BENCH, a collection of 20 NLP datasets with human annotations, and comprehensively evaluate 11 current LLMs, covering both open-weight and proprietary models, for their ability to replicate the annotations. Our evaluations show that each LLM exhibits a large variance across datasets in its correlation to human judgments. We conclude that LLMs are not yet ready to systematically replace human judges in NLP.

Fine-grained Controllable Text Generation through In-context Learning with Feedback

We present a method for rewriting an input sentence to match specific values of nontrivial linguistic features, such as dependency depth. In contrast to earlier work, our method uses in-context learning rather than finetuning, making it applicable in use cases where data is sparse. We show that our model performs accurate rewrites and matches the state of the art on rewriting sentences to a specified school grade level.