Jerome Feldman

Jerome Feldman
University of California, Berkeley | UCB

Ph.D. 1964 Carnegie-Mellon

About

199
Publications
17,856
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
6,914
Citations
Introduction
An updated summary of recent work has been published as: Feldman, J. (2022). Computation, perception, and mind. Behavioral and Brain Sciences, 45, E48.doi:10.1017/S0140525X21001886 A more readable, open access, version is : https://escholarship.org/uc/item/6cs78450
Additional affiliations
January 1988 - March 2014
University of California, Berkeley
Position
  • Professor (Full)
January 1986 - October 2015
University of California, Berkeley
Position
  • Professor
June 1974 - June 1988
University of Rochester
Position
  • Professor

Publications

Publications (199)
Article
Full-text available
Advances in behavioral and brain sciences have engendered wide ranging efforts to help understand consciousness. The target article suggests that abstract computational models (such as integrated information theory [IIT]) are ill-advised. This commentary broadens the discussion to include mysteries of subjective experience that are inconsistent wit...
Preprint
Full-text available
Subjective Experience (SE) is part of the ancient mind-body problem, which continues to be one of deepest mysteries of science. Despite major advances in many fields, there is still no plausible causal link between SE and its realization in the body. The core issue is the incompatibility of objective (3rd person) public science with subjective (1st...
Article
This paper describes the continuing goals and present status of the ICSI/UC Berkeley efforts on Embodied Construction Grammar (ECG). ECG is semantics-based formalism grounded in cognitive linguistics. ECG is the most explicitly inter-disciplinary of the construction grammars with deep links to computation, neuroscience, and cognitive science. Work...
Preprint
Full-text available
Abstract Subjective Experience (SE) is part of the ancient mind-body problem, which continues to be one of deepest mysteries of science. Despite major advances in many fields, there is still no plausible causal link between SE and its realization in the body. The core issue is the incompatibility of objective (3rd person) public science with subje...
Book
Full-text available
Welcome to the 18th IEEE International Conference on Cognitive Informatics and Cognitive Computing (ICCI*CC 2019), a flagship conference in this field, at Politecnico di Milano University, Italy. The growing field of CI and CC covers transdisciplinary areas such as brain-inspired intelligence, abstract intelligence, cognitive computing, neuroinform...
Article
Full-text available
Sweeney’s lament here was that words could not express the full complexity of his feelings and thoughts. The contested usage of technical terms is also endemic in discussions of two of the most profound scientific mysteries: the mind/body problem and the ultimate nature of physical reality. There are still scientific communities that deny the exist...
Preprint
Full-text available
The ancient mind/body problem continues to be one of deepest mysteries of science and of the human spirit. Despite major advances in many fields, there is still no plausible link between subjective experience (qualia) and its realization in the body. This paper outlines some of the elements of a rigorous science of mind (SoM) - key ideas include sc...
Book
Full-text available
IEEE ICCI*CC’18 organized at UC Berkeley from July 16-18, 32018 focuses on the theme of Cognitive Machine Learning, Brain-Inspired Systems and Cognitive Robotics towards developing cognitive systems in cognitive informatics, cognitive computing, computational intelligence, and neural informatics. Cognitive Informatics (CI) is a transdisciplinary...
Article
The Association for the Advancement of Artificial Intelligence, in cooperation with Stanford University's Department of Computer Science, presented the 2017 Spring Symposium Series, held Monday through Wednesday, March 27–29, 2017 on the campus of Stanford University. The eight symposia held were Artificial Intelligence for the Social Good (SS-17-0...
Article
Full-text available
There remains considerable controversy about how the brain operates. This review focuses on brain activity rather than just structure and on concepts of action and actionability rather than truth conditions. Neural Communication is reviewed as a crucial aspect of neural encoding. Consequently, logical inference is superseded by neural simulation. S...
Article
Full-text available
Actions may not proceed as planned; they may be interrupted, resumed or overridden. This is a challenge to handle in a natural language understanding system. We describe extensions to an existing implementation for the control of autonomous systems by natural language, to enable such systems to handle incoming language requests regarding actions. L...
Article
Full-text available
Science is a crowning glory of the human spirit and its applications remain our best hope for social progress. But there are limitations to current science and perhaps to any science. The general mind-body problem is known to be intractable and currently mysterious. This is one of many deep problems that are universally agreed to be beyond the curr...
Article
Full-text available
We develop a natural language interface for human robot interaction that implements reasoning about deep semantics in natural language. To realize the required deep analysis, we employ methods from cognitive linguistics, namely the modular and compositional framework of Embodied Construction Grammar (ECG) [Feldman, 2009]. Using ECG, robots are able...
Article
Full-text available
The famous Neural Binding Problem (NBP) comprises at least four distinct problems with different computational and neural requirements. This review discusses the current state of work on General Coordination, Visual Feature-Binding, Variable Binding, and the Subjective Unity of Perception. There is significant continuing progress, partially masked...
Conference Paper
Constraint satisfaction has been central to the ICSI/UC Berkeley Neural Theory of Language (NTL) project, but this aspect has not previously been emphasized. The ECG Analysis program combines constraints from several aspects of the formalism: deep semantic schemas, embodied constructions and ontological knowledge. In this chapter we focus on some a...
Article
Full-text available
The biology of cognition and language understanding is several decades behind the levels achieved in perception and motor control. This arises partly from a lack of animal models for language, but there are also conceptual issues. Even now, much of the work on language and cognition is formulated in terms that make no contact with biology and are o...
Chapter
Recent developments in neuroscience and behavioural sciences suggest approaching language as a cornerstone of Unified Cognitive Science. One such integrative effort is the Neural Theory of Language (NTL) project, which studies language learning and use as an embodied neural system using a wide range of analytical, experimental, and modelling techni...
Article
Full-text available
This review brings together two fundamental, but unreconciled, aspects of human language: embodiment and compositionality. One major scientific advance in recent decades has been Embodiment - the realization that scientific understanding of mind and language entails detailed modeling of the human brain and how it evolved to control a physical body...
Chapter
Full-text available
Communities of Learning Practice is an innovative paradigm focused on providing appropriate technological support to both formal and especially informal learning groups who are chiefly formed by non-technical people and who lack of the necessary resources to acquire such systems. Typically, students who are often separated by geography and/or time...
Chapter
Full-text available
Computer Supported Cooperative Work (CSCW), Democratic Group Production and e-Learning have overlapping but not identical design and use profiles. We suggest that democratically organized CSCW systems provide an ideal platform for organizing and encouraging group interactions in e-Learning and other democratic group production situations where phys...
Article
Full-text available
Neural spikes are an evolutionarily ancient innovation that remains nature's unique mechanism for rapid, long distance information transfer. It is now known that neural spikes sub serve a wide variety of functions and essentially all of the basic questions about the communication role of spikes have been answered. Current efforts focus on the neura...
Chapter
A coherent and plausible picture of human concept learning is arising from combining biological, behavioral, computational, and linguistic insights. This account is similar in form to the solution to another biological question-a question for which the answer is now understood in great detail. That is the question of immunology. Animal immune syste...
Article
Full-text available
To perform automatic, unconscious inference, the human brain must solve the binding problem by correctly grouping properties with objects. Temporal binding models like SHRUTI already suggest much of how this might be done in a connectionist and localist way by using temporal synchrony. We propose a set of alternatives to temporal synchrony mechanis...
Article
Full-text available
i L i i L-i IL-i i L The problem of inferring a grammar for a set of symbol strings is considered and a number of new decidability results obtained. Several notions of grammatical complexity and their properties are studied. The question of learning the least complex grammar for a
Conference Paper
Full-text available
Computer supported cooperative work (CSCW), and e-learning have overlapping, but not identical design and use profiles. We suggest that democratically organized CSCW systems provide an ideal platform for organizing and encouraging group interactions in e-learning situations where physical meetings are not feasible. Few attempts at democratic CSCW s...
Conference Paper
Full-text available
Communities of learning practice is an innovative paradigm focused on providing appropriate technological support to both formal and especially informal learning groups who are chiefly formed by non-technical people and who lack of the necessary resources to acquire such systems. Typically, students who are often separated by geography and/or time...
Article
Full-text available
Does sentence comprehension related to faces modulate activity in the fusiform face area (FFA) and does sentence comprehension related to places modulate activity in the parahippocampal place area (PPA)? We investigated this question in an fMRI experiment. Participants listened to sentences describing faces, places, or objects, with the latter serv...
Article
Full-text available
This paper reports on the experience of using an innovative technology-enhanced learning tool to support a real community of formal learning practice. First, we introduce the underlying groupware platform, called CoPE, that provides the essential functional support for democratic groupware. Then, we describe the main guidelines for the requirements...
Chapter
Science must eventually explain how the properties of the brain determine cognitive behavior. Connectionist approaches to cognitive science are based on the belief that our present understanding of neural and brain physiology should guide cognitive research.
Article
On Intelligence by Jeff Hawkins with Sandra Blakeslee has been inspirational for non- scientists as well as some of our most distinguished biologists, as can be seen from the web site (http://www.onintelligence.org). The book is engagingly written as a first person memoir of one computer engineer's search for enlightenment on how human intelligence...
Article
Various disciplines have examined the many phenomena of metacognition and have produced numerous results, both positive and negative. I discuss some of these aspects of cognition about cognition and the results concerning them from the point of view ...
Article
Full-text available
The Neural Theory of Language project aims to build structured connectionist models of language and cognition consistent with constraints from all domains and at all levels. These constraints include recent experimental evidence that details of neural computation and brain architecture play a crucial role in language processing. We focus in this pa...
Article
Full-text available
The Communities of Practice Environment (CoPE) project is an exploration of the potential for extending the social processes of the open source software movement to a wide range of other cooperative activities. The general ideas are best expressed in Steven Weber's book: The Success Of Open Source, Harvard U. Press, 2004 (1). Obviously enough, coll...
Article
Full-text available
1 The talk will describe an ongoing project (modestly named the Neural Theory of Language) that is attempting to model language behavior in a way that is both neurally plausible and computationally practical. The cornerstone of the effort is a formalism called Embodied Construction Grammar (ECG). I will describe the formalism, a robust semantic par...
Article
Full-text available
Imagine that you are dining with a group of people and one of them asks you to pass the salt. Ignoring for now any social or emotional issues, what does it mean for you to carry out the request? You must locate the salt container, determine which grip to use, reach out and grasp the container, and move it to a location near the requester. If the sa...
Article
Full-text available
Mental Space Theory (Fauconnier, 1985) en-compasses a wide variety of complex linguis-tics phenomena that are largely ignored in to-day's natural language processing systems. These phenomena include conditionals (e.g. If sentences), embedded discourse, and other natural language utterances whose interpreta-tion depends on cognitive partitioning of...
Article
Research in cognitive linguistics has yielded valuable semantic and pragmatic insights that should be incorporated by language understanding systems. These insights have not been expressed in a rigorous enough form to be implemented.
Conference Paper
Full-text available
The requirement to train large neural networks quickly has prompted the design of a new massively parallel supercomputer using custom VLSI. This design features 128 processing nodes, communicating over a mesh network connected directly to the processor chip. Studies show peak performance in the range of 160 billion arithmetic operations per second....
Article
Full-text available
. Wepresent a connectionist realization of parameterized schemas that can model high-level sensory-motor processes and be a candidate representation for implementing reactivebehaviors. The connectionist realization involves a number of ideas including the use of focal-clusters and feedback loops to control a distributed process without a central co...
Article
Full-text available
The transition to the next millennium gives us an opportunity to reflect on the past and project the future. In this spirit, we have asked a set of distinguished scholars and practitioners who were involved in AI's formative stages to describe the most notable trend or controversy (or nontrend or noncontroversy) during AI's development. The respons...
Article
On the whole, I quite like both the target book (RI) and the review by Matthew Rispoli, but there are some additional observations that might be of interest to the readers of this journal. The whole concept of ‘truth through disputation’ is alien to my scientific tradition and I agree with Rispoli that RI is not helped by the polemical tone. There...
Conference Paper
Direct connnectionist modeling of higher cognitive functions, such as language understanding, is impractical. This chapter describes a principled multi-layer architecture that supports AI style computational modeling while preserving the biological plausibility of structured connectionist models. As an example, the connectionist realization of Baye...
Article
Full-text available
This paper presents an implemented computational model of lexical development for the case of action verbs. A simulated agent is trained by an informant giving labels to the agent's actions (here hand motions) and the system learns to both label and carry out similar actions. Computationally, the system employs a novel form of active representation...
Article
This paper describes an implemented computational model of lexical development for the case of action verbs.
Conference Paper
Formal studies of learning abstract grammars have, for decades, yielded deep results, some practical applications and quite a lot of fun. But no one believes that children learn the grammar of their native language independent of meaning (semantics) and use (pragmatics). Recent results suggest that is now possible, although still very difficult, to...
Article
Full-text available
ions to Guide Strongly Typed Library Design B. Gomes, D. Stoutamire, B. Weissman and J. Feldman TR-97-061 December 1997 Abstract This report addresses typing problems that arise when modelling simple mathematical entities in strongly typed languages such as Sather, which are eliminated by a proper distinction between mutable and immutable abstracti...
Article
Full-text available
This report proposes a massively parallel computer, the Connectionist Network Supercomputer(CNS-1), which leverages off these fields. By targeting the computer to connectionist networks and related applications, we can focus on custom chip design and efficient software to achieve performance goals which challenge the best that industry has to offer...
Article
Full-text available
pSather is a parallel extension of Sather aimed at shared memory parallel architectures. A prototype of the language is currently being implemented on a Sequent Symmetry and on SUN Sparc-Stations. pSather monitors are one of the basic new features introduced in the language to deal with parallelism. The current design is presented and discussed in...
Article
Full-text available
pSather is a parallel extension of the existing object-oriented language Sather. It offers a shared-memory programming model which integrates both control- and dataparallel extensions. This integration increases the flexibility of the language to express different algorithms and data structures, especially on distributed-memory machines (e.g. CM-5)...
Article
This report proposes a massively parallel computer, the Connectionist Network Supercomputer(CNS-1), which leverages off these fields. By targeting the computer to connectionist networks and related applications, we can focus on custom chip design and efficient software to achieve performance goals which challenge the best that industry has to offer...
Article
Embodiment, the explicit dependence of cognition on the properties of the human body, is the foundation of contemporary cognitive science. Ballard et al.'s target article makes an important contribution to the embodiment story by suggesting how limitations on neural binding ability lead to deictic strategies for many tasks. It also exploits th...
Article
The radical empiricist theory of the Quartz & Sejnowski target article would result in a brain that could not act. The attempt to bolster this position with computational arguments is misleading and often just wrong. Fortunately, other efforts are making progress in linking neural and cognitive development.
Article
Full-text available
This chapter describes several probabilistic techniques for representing, recognizing, and generating spatiotemporal configuration sequences. We first describe how such techniques can be used to visually track and recognize lip movements to augment a speech recognition system. We then demonstrate additional techniques that can be used to animate vi...
Article
Full-text available
This paper describes an effort at UC Berkeley and the International Computer Science Institute to develop a super-computer for artificial neural network applications. We describe our applications targets, machine goals, and the system architecture. 1 Introduction The goals of the Connectionist Network Supercomputer-1 (CNS-1) project are to produce...
Chapter
Parallel Sather* (pSather [1]) is a parallel object-oriented language with a shared address space programming model, independent of the underlying multiprocessor architecture. To account for distributed-memory machines, pSather provides a cluster model in which processors are grouped in clusters. Associated with a cluster is a part of the address s...
Article
this article will focus on programming languages and their role in various HPC domains. The most advanced of these is scientific computation where various FORTRAN systems have yielded very impressive performance on Grand Challenge and related numerical calculations. Even in this domain, there is now widespread understanding that new languages and t...
Article
Full-text available
During the contract period the contractor: (1) Explored the requirements for two new applications in order to improve the capabilities for CNS-1 computation beyond the speech tasks that we have promised for the ONR-URI; (2) Developed library routines to support the general types of computation suggested by the tasks above as well as other tasks we...
Article
Full-text available
During the contract period we: (1) Considered connectionist computation, particularly asit has been used for applications in speech recognition, and developed perspectives on machine requiremebts for this task. (2) Analyzed the computational requirements for image understanding tasks, particularly in the context of the developing Arpa Image Underst...
Article
Full-text available
The L 0 project at ICSI and UC Berkeley attempts to combine not only vision and natural language modelling, but also learning. The original task was put forward in (Feldman et al. 1990a) as a touchstone task for AI and cognitive science. The task is to build a system that can learn the appropriate fragment of any natural language from sentence-pict...
Article
Full-text available
This paper describes an effort at UC Berkeley and the International Computer Science Institute to develop a supercomputer for artificial neural network applications. Our perspective has been strongly influenced by earlier experiences with the construction and use of a simpler machine. In particular, we have observed Amdahl's Law in action in our de...
Conference Paper
The dusty deck problem is the problem of how to automatically transform large existing serial/vector codes into a form suitable for efficient execution on new parallel architectures. Kernel-control parallel (KCP) methods provide a promising ...
Article
Full-text available
Connectionist learning models have had considerable empirical success, but it is hard to characterize exactly what they learn. The learning of finite-state languages (FSL) from example strings is a domain which has been extensively studied and might provide an opportunity to help understand connectionist learning. A major problem is that traditiona...
Article
To perceive a stable world, one must somehow be able to relate visual information from successive fixations. Little is known, however, about the nature of the integrative process. By using a task which requires the integration of spatial position information from different fixations, it is demonstrated that visual information from previous fixation...
Chapter
It has been clear for many years that parallelism is the cornerstone of future computing. What has been much less clear is how to achieve high degrees of parallelism in a practical way. In the standard formulation there is an inherent dilemma facing the designers and users of highly parallel systems. It is relatively easy and efficient to build loo...
Article
Full-text available
Cognitive Science, whose genesis was interdisciplinary, shows signs of reverting to a disjoint collection of fields. This paper presents a compact, theory-free task that inherently requires an integrated solution. The basic problem is learning a subset of an arbitrary natural language from picture-sentence pairs. We describe a very specific instanc...
Article
Artificial intelligence and, more recently, neural networks have been claimed to yield revolutionary advances in manufacturing. József Hatvany had the clearest view of the advantages and problems with AI in manufacturing. The situation with neural networks is similar and can benefit from an extension of Hatvany's analysis.