arXiv subjects

Santiago Ontanon

Publications and source records attributed to Santiago Ontanon.

13 recordsLinked to original sources

arXiv2024

Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context

In this report, we introduce the Gemini 1.5 family of models, representing the next generation of highly compute-efficient multimodal models capable of recalling and reasoning over fine-grained information from millions of tokens of context, including multiple long documents and hours of video and audio. The family includes two new models: (1) an updated Gemini 1.5 Pro, which exceeds the February version on the great majority of capabilities and benchmarks; (2) Gemini 1.5 Flash, a more lightweight variant designed for efficiency with minimal regression in quality. Gemini 1.5 models achieve near-perfect recall on long-context retrieval tasks across modalities, improve the state-of-the-art in long-document QA, long-video QA and long-context ASR, and match or surpass Gemini 1.0 Ultra's state-of-the-art performance across a broad set of benchmarks. Studying the limits of Gemini 1.5's long-context ability, we find continued improvement in next-token prediction and near-perfect retrieval (>99%) up to at least 10M tokens, a generational leap over existing models such as Claude 3.0 (200k) and GPT-4 Turbo (128k). Finally, we highlight real-world use cases, such as Gemini 1.5 collaborating with professionals on completing their tasks achieving 26 to 75% time savings across 10 different job categories, as well as surprising new capabilities of large language models at the frontier; when given a grammar manual for Kalamang, a language with fewer than 200 speakers worldwide, the model learns to translate English to Kalamang at a similar level to a person who learned from the same content.

cs.CL↗

arXiv2024

Michelangelo: Long Context Evaluations Beyond Haystacks via Latent Structure Queries

We introduce Michelangelo: a minimal, synthetic, and unleaked long-context reasoning evaluation for large language models which is also easy to automatically score. This evaluation is derived via a novel, unifying framework for evaluations over arbitrarily long contexts which measure the model's ability to do more than retrieve a single piece of information from its context. The central idea of the Latent Structure Queries framework (LSQ) is to construct tasks which require a model to ``chisel away'' the irrelevant information in the context, revealing a latent structure in the context. To verify a model's understanding of this latent structure, we query the model for details of the structure. Using LSQ, we produce three diagnostic long-context evaluations across code and natural-language domains intended to provide a stronger signal of long-context language model capabilities. We perform evaluations on several state-of-the-art models and demonstrate both that a) the proposed evaluations are high-signal and b) that there is significant room for improvement in synthesizing long-context information.

cs.CL↗

arXiv2024

Functional Interpolation for Relative Positions Improves Long Context Transformers

Preventing the performance decay of Transformers on inputs longer than those used for training has been an important challenge in extending the context length of these models. Though the Transformer architecture has fundamentally no limits on the input sequence lengths it can process, the choice of position encoding used during training can limit the performance of these models on longer inputs. We propose a novel functional relative position encoding with progressive interpolation, FIRE, to improve Transformer generalization to longer contexts. We theoretically prove that this can represent some of the popular relative position encodings, such as T5's RPE, Alibi, and Kerple. We next empirically show that FIRE models have better generalization to longer contexts on both zero-shot language modeling and long text benchmarks.

cs.LG↗

arXiv2023

Multi-Task End-to-End Training Improves Conversational Recommendation

In this paper, we analyze the performance of a multitask end-to-end transformer model on the task of conversational recommendations, which aim to provide recommendations based on a user's explicit preferences expressed in dialogue. While previous works in this area adopt complex multi-component approaches where the dialogue management and entity recommendation tasks are handled by separate components, we show that a unified transformer model, based on the T5 text-to-text transformer model, can perform competitively in both recommending relevant items and generating conversation dialogue. We fine-tune our model on the ReDIAL conversational movie recommendation dataset, and create additional training tasks derived from MovieLens (such as the prediction of movie attributes and related movies based on an input movie), in a multitask learning setting. Using a series of probe studies, we demonstrate that the learned knowledge in the additional tasks is transferred to the conversational setting, where each task leads to a 9%-52% increase in its related probe score.

cs.CL↗

arXiv2022

FNet: Mixing Tokens with Fourier Transforms

We show that Transformer encoder architectures can be sped up, with limited accuracy costs, by replacing the self-attention sublayers with simple linear transformations that "mix" input tokens. These linear mixers, along with standard nonlinearities in feed-forward layers, prove competent at modeling semantic relationships in several text classification tasks. Most surprisingly, we find that replacing the self-attention sublayer in a Transformer encoder with a standard, unparameterized Fourier Transform achieves 92-97% of the accuracy of BERT counterparts on the GLUE benchmark, but trains 80% faster on GPUs and 70% faster on TPUs at standard 512 input lengths. At longer input lengths, our FNet model is significantly faster: when compared to the "efficient" Transformers on the Long Range Arena benchmark, FNet matches the accuracy of the most accurate models, while outpacing the fastest models across all sequence lengths on GPUs (and across relatively shorter lengths on TPUs). Finally, FNet has a light memory footprint and is particularly efficient at smaller model sizes; for a fixed speed and accuracy budget, small FNet models outperform Transformer counterparts.

cs.CL↗

arXiv2022

LongT5: Efficient Text-To-Text Transformer for Long Sequences

Recent work has shown that either (1) increasing the input length or (2) increasing model size can improve the performance of Transformer-based neural models. In this paper, we present a new model, called LongT5, with which we explore the effects of scaling both the input length and model size at the same time. Specifically, we integrated attention ideas from long-input transformers (ETC), and adopted pre-training strategies from summarization pre-training (PEGASUS) into the scalable T5 architecture. The result is a new attention mechanism we call {\em Transient Global} (TGlobal), which mimics ETC's local/global attention mechanism, but without requiring additional side-inputs. We are able to achieve state-of-the-art results on several summarization tasks and outperform the original T5 models on question answering tasks.

cs.CL↗

arXiv2022

LogicInference: A New Dataset for Teaching Logical Inference to seq2seq Models

Machine learning models such as Transformers or LSTMs struggle with tasks that are compositional in nature such as those involving reasoning/inference. Although many datasets exist to evaluate compositional generalization, when it comes to evaluating inference abilities, options are more limited. This paper presents LogicInference, a new dataset to evaluate the ability of models to perform logical inference. The dataset focuses on inference using propositional logic and a small subset of first-order logic, represented both in semi-formal logical notation, as well as in natural language. We also report initial results using a collection of machine learning models to establish an initial baseline in this dataset.

cs.AI↗

arXiv2021

Rotating Boson Stars Using Finite Differences and Global Newton Methods

We study Rotating Boson Star initial data for Numerical Relativity as previously considered by Yoshida and Eriguchi, Lai (arXiv:gr-qc/0410040v2), and Grandclement, Somé and Gourgoulhon (arXiv:1405.4837v3). We use a 3 + 1 decomposition as presented by Gourgoulhon (arXiv:1003.5015v2) and Alcubierre, adapted to an axisymmetric quasi-isotropic spacetime with added regularization at the axis following work by Ruíz, Alcubierre and Núñez (arXiv:0706.0923v2) and Torres. The Einstein-Klein-Gordon equations result in a system of six-coupled, elliptic, nonlinear equations with an added unknown for the scalar field's frequency $ω$. Utilizing a Cartesian two-dimensional grid, finite differences, Global Newton Methods adapted from Deuflhard, the sparse direct linear solver PARDISO, and properly constraining all variables generates data sets for rotation azimuthal integers $l \in [0, 6]$. Our numerical implementation, published in GitHub, is shown to correctly converge both with respect to the resolution size and boundary extension (fourth-order and third-order, respectively). Thus, global parameters such as the Komar masses and angular momenta can be precisely calculated to characterize these spacetimes. Furthermore, analyzing the full family at fixed rotation integer produces maximum masses and minimum frequencies. These coincide with previous results in literature for $l \in [0,2]$ and are new for $l > 2$. In particular, the study of high-amplitude and localized scalar fields in axial symmetry is revealed to be only possible by adding the sixth regularization variable.

gr-qc↗

arXiv2021

Big Bird: Transformers for Longer Sequences

Transformers-based models, such as BERT, have been one of the most successful deep learning models for NLP. Unfortunately, one of their core limitations is the quadratic dependency (mainly in terms of memory) on the sequence length due to their full attention mechanism. To remedy this, we propose, BigBird, a sparse attention mechanism that reduces this quadratic dependency to linear. We show that BigBird is a universal approximator of sequence functions and is Turing complete, thereby preserving these properties of the quadratic, full attention model. Along the way, our theoretical analysis reveals some of the benefits of having $O(1)$ global tokens (such as CLS), that attend to the entire sequence as part of the sparse attention mechanism. The proposed sparse attention can handle sequences of length up to 8x of what was previously possible using similar hardware. As a consequence of the capability to handle longer context, BigBird drastically improves performance on various NLP tasks such as question answering and summarization. We also propose novel applications to genomics data.

cs.LG↗

arXiv2020

ETC: Encoding Long and Structured Inputs in Transformers

Transformer models have advanced the state of the art in many Natural Language Processing (NLP) tasks. In this paper, we present a new Transformer architecture, Extended Transformer Construction (ETC), that addresses two key challenges of standard Transformer architectures, namely scaling input length and encoding structured inputs. To scale attention to longer inputs, we introduce a novel global-local attention mechanism between global tokens and regular input tokens. We also show that combining global-local attention with relative position encodings and a Contrastive Predictive Coding (CPC) pre-training objective allows ETC to encode structured inputs. We achieve state-of-the-art results on four natural language datasets requiring long and/or structured inputs.

cs.LG↗

arXiv2014

A semantic network-based evolutionary algorithm for computational creativity

We introduce a novel evolutionary algorithm (EA) with a semantic network-based representation. For enabling this, we establish new formulations of EA variation operators, crossover and mutation, that we adapt to work on semantic networks. The algorithm employs commonsense reasoning to ensure all operations preserve the meaningfulness of the networks, using ConceptNet and WordNet knowledge bases. The algorithm can be interpreted as a novel memetic algorithm (MA), given that (1) individuals represent pieces of information that undergo evolution, as in the original sense of memetics as it was introduced by Dawkins; and (2) this is different from existing MA, where the word "memetic" has been used as a synonym for local refinement after global optimization. For evaluating the approach, we introduce an analogical similarity-based fitness measure that is computed through structure mapping. This setup enables the open-ended generation of networks analogous to a given base network.

cs.NE↗

arXiv2012

Experiments with Game Tree Search in Real-Time Strategy Games

Game tree search algorithms such as minimax have been used with enormous success in turn-based adversarial games such as Chess or Checkers. However, such algorithms cannot be directly applied to real-time strategy (RTS) games because a number of reasons. For example, minimax assumes a turn-taking game mechanics, not present in RTS games. In this paper we present RTMM, a real-time variant of the standard minimax algorithm, and discuss its applicability in the context of RTS games. We discuss its strengths and weaknesses, and evaluate it in two real-time games.

cs.AI↗

arXiv2012

Automated Generation of Cross-Domain Analogies via Evolutionary Computation

Analogy plays an important role in creativity, and is extensively used in science as well as art. In this paper we introduce a technique for the automated generation of cross-domain analogies based on a novel evolutionary algorithm (EA). Unlike existing work in computational analogy-making restricted to creating analogies between two given cases, our approach, for a given case, is capable of creating an analogy along with the novel analogous case itself. Our algorithm is based on the concept of "memes", which are units of culture, or knowledge, undergoing variation and selection under a fitness measure, and represents evolving pieces of knowledge as semantic networks. Using a fitness function based on Gentner's structure mapping theory of analogies, we demonstrate the feasibility of spontaneously generating semantic networks that are analogous to a given base network.

cs.NE↗