From decoding to meta-generation: Inference-time algorithms for large language models
One of the most striking findings in modern research on large language models (LLMs) is
that scaling up compute during training leads to better results. However, less attention has …
that scaling up compute during training leads to better results. However, less attention has …
K⁎: A heuristic search algorithm for finding the k shortest paths
We present a directed search algorithm, called K⁎, for finding the k shortest paths between a
designated pair of vertices in a given directed weighted graph. K⁎ has two advantages …
designated pair of vertices in a given directed weighted graph. K⁎ has two advantages …
If the current clique algorithms are optimal, so is Valiant's parser
The CFG recognition problem is as follows: given a context-free grammar G and a string w of
length n, decide whether w can be obtained from G. This is the most basic parsing question …
length n, decide whether w can be obtained from G. This is the most basic parsing question …
Imitation learning of agenda-based semantic parsers
Semantic parsers conventionally construct logical forms bottom-up in a fixed order, resulting
in the generation of many extraneous partial logical forms. In this paper, we combine ideas …
in the generation of many extraneous partial logical forms. In this paper, we combine ideas …
Dyna: Extending datalog for modern AI
J Eisner, NW Filardo - International Datalog 2.0 Workshop, 2010 - Springer
Modern statistical AI systems are quite large and complex; this interferes with research,
development, and education. We point out that most of the computation involves database …
development, and education. We point out that most of the computation involves database …
[BUCH][B] Text analysis pipelines: towards ad-hoc large-scale text mining
H Wachsmuth - 2015 - books.google.com
This monograph proposes a comprehensive and fully automatic approach to designing text
analysis pipelines for arbitrary information needs that are optimal in terms of run-time …
analysis pipelines for arbitrary information needs that are optimal in terms of run-time …
Global neural CCG parsing with optimality guarantees
We introduce the first global recursive neural parsing model with optimality guarantees
during decoding. To support global features, we give up dynamic programs and instead …
during decoding. To support global features, we give up dynamic programs and instead …
-best enumeration
D Eppstein - arxiv preprint arxiv:1412.5075, 2014 - arxiv.org
arxiv:1412.5075v1 [cs.DS] 16 Dec 2014 Page 1 arxiv:1412.5075v1 [cs.DS] 16 Dec 2014 K-Best
Enumeration David Eppstein∗ Abstract We survey k-best enumeration problems and the …
Enumeration David Eppstein∗ Abstract We survey k-best enumeration problems and the …
On Finding the -best Non-projective Dependency Trees
The connection between the maximum spanning tree in a directed graph and the best
dependency tree of a sentence has been exploited by the NLP community. However, for …
dependency tree of a sentence has been exploited by the NLP community. However, for …
[PDF][PDF] Efficient inference through cascades of weighted tree transducers
Weighted tree transducers have been proposed as useful formal models for representing
syntactic natural language processing applications, but there has been little description of …
syntactic natural language processing applications, but there has been little description of …