Link to paper
The full paper is available here.
You can also find the paper on PapersWithCode here.
Abstract
- Presents Integrated Information Theory (IIT) 4.0
- Aims to explain properties of experience in physical terms
- Identifies essential properties of experience and expresses them mathematically
- Can be applied to any system to determine if it is conscious
- Makes testable predictions and allows inferences and extrapolations
- Includes more accurate translation of axioms, a measure of intrinsic information, and an assessment of causal relations
- Unfolds a system’s cause-effect power to explain the quality of experience
Paper Content
Introduction
- Consciousness is subjective and IIT aims to explain it in physical terms
- IIT starts with the existence of an experience, which is immediate and irrefutable
- IIT has five axioms of phenomenal existence: intrincicality, information, integration, exclusion, and composition
- Physical existence is assessed operationally from within consciousness
- IIT proposes a fundamental explanatory identity: an experience is identical to the cause-effect structure unfolded from a maximal substrate
- IIT has a mathematical framework to evaluate self-consistency and make predictions
- IIT has been refined and extended over time
- IIT has a book and wiki to explain motivation, axioms, postulates, and assumptions
Postulates of physical existence
- Realism: Assumption of a world that persists independently of one’s experience
- Physicalism: Something must have the power to take and make a difference in a reliable way to be granted physical existence
- Operational Reductionism: Establishing what exists in physical terms by starting from the smallest units
- Intrinsicality: Substrate of consciousness must have intrinsic cause-effect power
- Information: Substrate of consciousness must have specific cause-effect power
- Integration: Substrate of consciousness must have unitary cause-effect power
- Exclusion: Substrate of consciousness must have definite cause-effect power
- Composition: Substrate of consciousness must have structured cause-effect power
The explanatory identity between experiences and φ-structures
- IIT proposes an explanatory identity that states that the physical properties of a complex can explain the properties of an experience.
- The identity is not meant to be a correspondence between two separate things, but an explanation of the intrinsic feeling of an experience in terms of cause-effect power.
Box 1. methodological guidelines of iit
- Explanations should be broad, unified, precise, coherent, consistent with overall understanding, and simpler than alternatives.
Overview of iit’s framework
- IIT formalism identifies sets of units that fulfill properties of a substrate of consciousness
- IIT determines a maximal cause-effect state based on intrinsic information of the system
- IIT determines a maximal substrate based on integrated information of the maximal cause-effect state
- IIT unfolds the cause-effect structure specified by a maximal substrate in its current state
- IIT characterizes Φ-structures based on measures that satisfy the postulates uniquely
Substrates, transition probabilities, and cause-effect power
- IIT takes physical existence as synonymous with having cause-effect power
- A substrate is operationally defined by its potential interactions
- A substrate is a stochastic system of interacting units with a finite state space
- The transition probability function of a substrate is determined by a transition probability matrix
- IIT does not require intrinsic physical properties or laws of nature
- IIT is applied to simulated networks to illustrate cause-effect power
Implementing the postulates
- Goal is to evaluate whether a hypothetical substrate satisfies postulates of IIT
- Must verify if system has cause-effect power that is intrinsic, specific, integrated, definite, and structured
- Existence requires capacity to take and make a difference
- Intrinsicality requires system exerts cause-effect power within itself
- Information postulate requires system’s cause-effect power be specific
- Integration postulate requires system specify maximal cause-effect state in irreducible way
- Exclusion postulate requires definite set of units, updates, and states
- Composition postulate requires evaluating cause-effect structure of subsets and their causal relations
- Formal account of IIT analysis includes determining cause-effect state, evaluating system integrated information, and searching for maxima of integrated information
- Second part of formal account includes identifying causal distinctions and relations, yielding system’s cause-effect structure and structured information
Box 2. ontological principles of iit
- To exist requires being able to take and make a difference
- Eleatic principle states that entities are nothing else but power
- Buddhist philosopher Dharmakīrti states that whatever has causal powers, that really exists
- Principle of maximal existence states that what exists is what exists the most
- Principle of minimal existence states that nothing exists more than the least it exists
- Degree of irreducibility should be assessed over the partition across which it is least irreducible
Identifying substrates of consciousness through existence, intrinsicality, information, integration, and exclusion
- Starting point is a substrate U in current state U = u with TPM T U
- Consider any subset s ⊆ u as a possible complex
- Refer to a set of units S ⊆ U as a candidate system
- Units W = U \ S are fixed in their current state w ∈ Ω W
- Obtain TPM T S of a candidate system S from its intrinsic transition probability function
- Intrinsic information ii c/e is a measure of the intrinsic cause/effect power exerted by a system S
- Integrated information ϕ s is a measure of the irreducibility of a cause-effect state
- Systems for which integrated information is maximal are called maximal substrates
Existence, intrinsicality, and information: determining the maximal cause-effect state of a candidate system
- A causal model is used to identify and quantify the cause-effect power of a system from its intrinsic perspective.
- Intrinsic information is the product of selectivity and informativeness.
- Intrinsic effect information is defined as the interventional conditional probability that the current state produces the effect state.
- The postulates of existence, intrinsicality, information, integration, and exclusion are used to identify substrates of consciousness.
- Existence requires that the substrate must have cause-effect power.
- Intrinsicality requires that a candidate substrate has cause-effect power over itself.
- Information requires that the candidate substrate selects a specific cause-effect state.
- Integration requires that the substrate specifies its cause-effect state irreducibly.
- Exclusion requires that the substrate of consciousness is definite.
- Intrinsic information is quantified in units of intrinsic bits.
- Intrinsic information captures the postulates of existence and intrinsicality.
- The principle of maximal existence determines which cause/effect state the system should select.
- Intrinsic information is formally equivalent to a measure of intrinsic difference.
Integration: determining the irreducibility of a candidate system
- Identified maximal cause-effect state of a candidate system
- Evaluated whether the system is irreducible
- Defined a set of directional system partitions to divide the system into parts
- Each part must contain at least one unit, no overlap between parts, and each unit must appear in exactly one part
- Partition removes the causal connections of that part with the rest of the system
- Defined a partitioned transition probability matrix
- Integrated effect information measures how much the partition reduces the probability of the system
- Integrated cause information is also defined
- System integrated information quantifies the extent to which the system has cause-effect power over itself
- System integrated information is a quantifier of irreducible existence
Exclusion: determining maximal substrates (complexes)
- Multiple candidate systems with overlapping units may have positive values of ϕ s
- Substrate of consciousness must be definite
- Principle of maximal existence: system with most existence is chosen
- Iterative search for system with maximum value of integrated information
- Substrates with lower ϕ s are excluded from existence
- Units must comply with postulates of IIT and be maximally irreducible