Search Results for author:
Found 67 papers, 47 papers with code
Self-Supervised Alignment with Mutual Information: Learning to Follow Principles without Preference Labels
On single-turn dialogue and summarization, a SAMI-trained mistral-7b outperforms the initial pretrained model, with win rates between 66% and 77%.
Procedural Dilemma Generation for Evaluating Moral Reasoning in Humans and Language Models
We collected moral permissibility and intention judgments from human participants for a subset of our items and compared these judgments to those from two language models (GPT-4 and Claude-2) across eight conditions.
Stream of Search (SoS): Learning to Search in Language
In this paper, we show how language models can be taught to search by representing the process of search in language, as a flattened string -- a stream of search (SoS).
STaR-GATE: Teaching Language Models to Ask Clarifying Questions
After two iterations of self-improvement, the Questioner asks better questions, allowing it to generate responses that are preferred over responses from the initial model on 72% of tasks.
Quiet-STaR: Language Models Can Teach Themselves to Think Before Speaking
Crucially, these improvements require no fine-tuning on these tasks.
pyvene: A Library for Understanding and Improving PyTorch Models via Interventions
Interventions on model-internal states are fundamental operations in many areas of AI, including model editing, steering, robustness, and interpretability.
Evaluating and Optimizing Educational Content with Large Language Model Judgments
We propose an alternative approach that uses Language Models (LMs) as educational experts to assess the impact of various instructions on learning outcomes.
Automated Statistical Model Discovery with Language Models
We evaluate our method in three common settings in probabilistic modeling: searching within a restricted space of models, searching over an open-ended space, and improving classic models under natural language constraints (e. g., this model should be interpretable to an ecologist).
A Reply to Makelov et al. (2023)'s "Interpretability Illusion" Arguments
We respond to the recent paper by Makelov et al. (2023), which reviews subspace interchange intervention methods like distributed alignment search (DAS; Geiger et al. 2023) and claims that these methods potentially cause "interpretability illusions".
Codebook Features: Sparse and Discrete Interpretability for Neural Networks
In this setting, our approach overcomes the superposition problem by assigning states to distinct codes, and we find that we can make the neural network behave as if it is in a different state by activating the code for that state.
Social Contract AI: Aligning AI Assistants with Implicit Group Norms
We explore the idea of aligning an AI assistant by inverting a model of users' (unknown) preferences from observed interactions.
CLEVRER-Humans: Describing Physical and Causal Events the Human Way
First, there is a lack of diversity in both event types and natural language descriptions; second, causal relationships based on manually-defined heuristics are different from human judgments.
Hypothesis Search: Inductive Reasoning with Language Models
Because of the prohibitive cost of generation with state-of-the-art LLMs, we consider a middle step to filter the set of hypotheses that will be implemented into programs: we either ask the LLM to summarize into a smaller set of hypotheses, or ask human annotators to select a subset of the hypotheses.
From Word Models to World Models: Translating from Natural Language to the Probabilistic Language of Thought
Our architecture integrates two computational tools that have not previously come together: we model thinking with probabilistic programs, an expressive representation for commonsense reasoning; and we model meaning construction with large language models (LLMs), which support broad-coverage translation from natural language utterances to code expressions in a probabilistic programming language.
Understanding Social Reasoning in Language Models with Language Models
Using our framework, we create a new social reasoning benchmark (BigToM) for LLMs which consists of 25 controls and 5, 000 model-written evaluations.
Just One Byte (per gradient): A Note on Low-Bandwidth Decentralized Language Model Finetuning Using Shared Randomness
Language model training in distributed settings is limited by the communication cost of gradient exchanges.
Certified Deductive Reasoning with Language Models
In experiments on PrOntoQA, ProofWriter and Syllogism Validity datasets, \textsc{LogicGuide} significantly improves the performance of GPT-3, GPT-3. 5 Turbo and LLaMA (accuracy gains up to 35\%), while drastically reducing \emph{content effects} -- the interference between unwanted prior assumptions and reasoning, which humans and language models suffer from.
Strategic Reasoning with Language Models
Existing approaches to solving strategic games rely on extensive training, yielding strategies that do not generalize to new scenarios or games without retraining.
Characterizing tradeoffs between teaching via language and demonstrations in multi-agent systems
Humans teach others about the world through language and demonstration.
Interpretability at Scale: Identifying Causal Mechanisms in Alpaca
With Boundless DAS, we discover that Alpaca does this by implementing a causal model with two interpretable boolean variables.
Overinformative Question Answering by Humans and Machines
When faced with a polar question, speakers often provide overinformative answers going beyond a simple "yes" or "no".
Bayesian Reinforcement Learning with Limited Cognitive Load
All biological and artificial agents must learn and make decisions given limits on their ability to process information.
Solving Math Word Problems by Combining Language Models With Symbolic Solvers
Automatically generating high-quality step-by-step solutions to math word problems has many applications in education.
Why think step by step? Reasoning emerges from the locality of experience
We investigate why and how chain-of-thought reasoning is useful in language models, testing the hypothesis that reasoning is effective when training data consists of overlapping local clusters of variables that influence each other strongly.
Finding Alignments Between Interpretable Causal Variables and Distributed Neural Representations
In DAS, we find the alignment between high-level and low-level models using gradient descent rather than conducting a brute-force search, and we allow individual neurons to play multiple distinct roles by analyzing representations in non-standard bases-distributed representations.
Parsel: Algorithmic Reasoning with Language Models by Composing Decompositions
Despite recent success in large language model (LLM) reasoning, LLMs struggle with hierarchical multi-step reasoning tasks like generating complex programs.
Ranked #8 on
Code Generation
on HumanEval
Geoclidean: Few-Shot Generalization in Euclidean Geometry
In contrast, low-level and high-level visual features from standard computer vision models pretrained on natural images do not support correct generalization.
Peano: Learning Formal Mathematical Reasoning
We explore this idea in a case study on 5 sections of beginning algebra on the Khan Academy platform.
On Rate-Distortion Theory in Capacity-Limited Cognition & Reinforcement Learning
Throughout the cognitive-science literature, there is widespread agreement that decision-making agents operating in the real world do so under limited information-processing capabilities and without access to unbounded cognitive or computational resources.
Psychologically-informed chain-of-thought prompts for metaphor understanding in large language models
Probabilistic models of language understanding are valuable tools for investigating human language use.
Color Overmodification Emerges from Data-Driven Learning and Pragmatic Reasoning
It seems likely that these patterns are shaped by the environment a speaker is exposed to in complex ways.
STaR: Bootstrapping Reasoning With Reasoning
We show that STaR significantly improves performance on multiple datasets compared to a model fine-tuned to directly predict final answers, and performs comparably to fine-tuning a 30$\times$ larger state-of-the-art language model on CommensenseQA.
Ranked #17 on
Common Sense Reasoning
on CommonsenseQA
Causal Distillation for Language Models
Distillation efforts have led to language models that are more compact and efficient without serious drops in performance.
Inducing Causal Structure for Interpretable Neural Networks
In IIT, we (1) align variables in a causal model (e. g., a deterministic program or Bayesian network) with representations in a neural model and (2) train the neural model to match the counterfactual behavior of the causal model on a base input when aligned representations in both models are set to be the value they would be for a source input.
Calibrate your listeners! Robust communication-based training for pragmatic speakers
We propose a method that uses a population of neural listeners to regularize speaker training.
Visual resemblance and communicative context constrain the emergence of graphical conventions
From photorealistic sketches to schematic diagrams, drawing provides a versatile medium for communicating about the visual world.
Learning to solve complex tasks by growing knowledge culturally across generations
The video game paradigm we pioneer here is thus a rich test bed for developing AI systems capable of acquiring and transmitting cultural knowledge.
Concadia: Towards Image-Based Text Generation with a Purpose
Current deep learning models often achieve excellent results on benchmark image-to-text datasets but fail to generate texts that are useful in practice.
From partners to populations: A hierarchical Bayesian account of coordination and convention
Languages are powerful solutions to coordination problems: they provide stable, shared expectations about how the words we say correspond to the beliefs and intentions in our heads.
Learning to refer informatively by amortizing pragmatic reasoning
A hallmark of human language is the ability to effectively and efficiently convey contextually relevant information.
Generalizing meanings from partners to populations: Hierarchical inference supports convention formation on networks
A key property of linguistic conventions is that they hold over an entire community of speakers, allowing us to communicate efficiently even with people we have never met before.
Characterizing the dynamics of learning in repeated reference games
The language we use over the course of conversation changes as we establish common ground and learn what our partner finds meaningful.
Continual adaptation for efficient machine communication
To communicate with new partners in new contexts, humans rapidly form new linguistic conventions.
Analyzing machine-learned representations: A natural language case study
Analyzing performance on these diagnostic tests indicates a lack of systematicity in the representations and decision rules, and reveals a set of heuristic strategies.
Learning to Explain: Answering Why-Questions via Rephrasing
We demonstrate that our strategy is sufficient to generate highly plausible explanations for general open-domain phenomena compared to other models trained on different datasets.
ShapeGlot: Learning Language for Shape Differentiation
We also find that these models are amenable to zero-shot transfer learning to novel object classes (e. g. transfer from training on chairs to testing on lamps), as well as to real-world images drawn from furniture catalogs.
Learning to Refer to 3D Objects with Natural Language
We further show that a neural speaker that is `listener-aware' --- that plans its utterances according to how an imagined listener would interpret its words in context --- produces more discriminative referring expressions than an `listener-unaware' speaker, as measured by human performance in identifying the correct object.
When redundancy is useful: A Bayesian approach to 'overinformative' referring expressions
Crucially, we relax the assumption that informativeness is computed with respect to a deterministic Boolean semantics, in favor of a non-deterministic continuous semantics.
Applying Probabilistic Programming to Affective Computing
Affective Computing is a rapidly growing field spurred by advancements in artificial intelligence, but often, held back by the inability to translate psychological theories of emotion into tractable computational models.
Joint Mapping and Calibration via Differentiable Sensor Fusion
Our algorithm outperforms our current production baseline based on k-means clustering.
Pyro: Deep Universal Probabilistic Programming
Pyro is a probabilistic programming language built on Python as a platform for developing advanced probabilistic models in AI research.
An Incremental Iterated Response Model of Pragmatics
Recent Iterated Response (IR) models of pragmatics conceptualize language use as a recursive process in which agents reason about each other to increase communicative efficiency.
The division of labor in communication: Speakers help listeners account for asymmetries in visual perspective
In Experiment 1, we manipulated the presence or absence of occlusions in a director-matcher task and found that speakers spontaneously produced more informative descriptions to account for "known unknowns" in their partner's private view.
Planning, Inference and Pragmatics in Sequential Language Games
We study sequential language games in which two players, each with private information, communicate to achieve a common goal.
Evaluating Compositionality in Sentence Embeddings
Further, we find that augmenting training with our dataset improves test performance on our dataset without loss of performance on the original training dataset.
DisSent: Sentence Representation Learning from Explicit Discourse Relations
Learning effective representations of sentences is one of the core missions of natural language understanding.
Learning Disentangled Representations with Semi-Supervised Deep Generative Models
We propose to learn such representations using model architectures that generalise from standard VAEs, employing a general graphical model structure in the encoder and decoder.
Colors in Context: A Pragmatic Neural Model for Grounded Language Understanding
We present a model of pragmatic referring expression interpretation in a grounded communication task (identifying colors from descriptions) that draws upon predictions from two recurrent neural network classifiers, a speaker and a listener, unified by a recursive pragmatic reasoning framework.
Inducing Interpretable Representations with Variational Autoencoders
We develop a framework for incorporating structured graphical models in the \emph{encoders} of variational autoencoders (VAEs) that allows us to induce interpretable representations through approximate variational inference.
Deep Amortized Inference for Probabilistic Programs
This paper proposes a system for amortized inference in PPLs.
The Language of Generalization
Language provides simple ways of communicating generalizable knowledge to each other (e. g., "Birds fly", "John hikes", "Fire makes smoke").
Learning to Generate Compositional Color Descriptions
The production of color language is essential for grounded language generation.
Neurally-Guided Procedural Models: Amortized Inference for Procedural Graphics Programs using Neural Networks
Probabilistic inference algorithms such as Sequential Monte Carlo (SMC) provide powerful tools for constraining procedural models in computer graphics, but they require many samples to produce desirable results.
Learning the Preferences of Ignorant, Inconsistent Agents
If we assume that choices are approximately optimal according to some utility function, we can treat preference inference as Bayesian inverse planning.
Coarse-to-Fine Sequential Monte Carlo for Probabilistic Programs
When models are expressed as probabilistic programs, the models themselves are highly structured objects that can be used to derive annealing sequences that are more sensitive to domain structure.
C3: Lightweight Incrementalized MCMC for Probabilistic Programs using Continuations and Callsite Caching
Lightweight, source-to-source transformation approaches to implementing MCMC for probabilistic programming languages are popular for their simplicity, support of existing deterministic code, and ability to execute on existing fast runtimes.
A Dynamic Programming Algorithm for Inference in Recursive Probabilistic Programs
This factored sum-product network makes (potentially cyclic) dependencies between subproblems explicit, and corresponds to a system of equations for the marginal distribution.
