Math 2.0

Resources / preprint

Limitations on Autoregressive Token Generation via Sequential Sampling 7D5MMP8XB5

URIm2://res/7D5MMP8XB5
Kindpreprint
CreatorsGal Arnon, Andrej Bogdanov, Paul Lou, Giulio Malavolta, Alon Rosen, Lorenzo Taschin
Date2026-10-10
Subjectsarxiv cs.IT, arxiv cs.AI, arxiv cs.CR, arxiv math.IT
Scopein

Expressions

e1: version, preprint

Languageen
Date2026-10-10
IdsarXiv 2610.14615

Items in this resource

The mentions of the inventory that name this resource. Each name links to the card of its inventory row.

ItemKindWhereSource
Small next-block pseudoentropy gap gives samplersstatementlemma 3.5arxiv
Batched table lookupstatementcorollary 2.15arxiv
Characterization of stateless sequential samplingstatementtheorem 3.3arxiv
45F9MJ9JFFstatementfact 2.13arxiv
Sample lower bound for estimating the state distributionstatementproposition 4.4arxiv
$(\ell, s, B)$-Sequential Sampling Algorithmdefinitiondefinition 3.1arxiv
Nonnegativity of KL divergence; see Theorem 2.6.3statementfact 2.4arxiv
Public-key steganographystatementcorollary 8.4arxiv
Data processing inequalitystatementfact 2.9arxiv
Separating insertional from next-block pseudoentropy gapsstatementcorollary 7.3arxiv
KL chain rule; Theorem 2.5.3statementfact 2.7arxiv
8DZC27GRFPstatementtheorem 1.3arxiv
Pseudorandom generationstatementcorollary 8.2arxiv
Insertional decompositiondefinitiondefinition 5.2arxiv
Distinct one-symbol insertionsstatementlemma 5.8arxiv
Small insertional pseudoentropy gives conditional samplersstatementlemma 5.11arxiv
Laplace estimatordefinitiondefinition 2.10arxiv
Markov chain; see Chapter 4definitiondefinition 2.6arxiv
One-way functions against circuitsdefinitiondefinition 2.18arxiv
Batched Laplace sampling from selected samplesstatementcorollary 2.17arxiv
General consequence for samplers with logarithmic statestatementcorollary 8.1arxiv
G4F5GD5MXHstatementtheorem 1.2arxiv
Insertional Pseudoentropydefinitiondefinition 5.3arxiv
Characterizing conditional pseudoentropy, nonuniform settingstatementtheorem 2.23arxiv
GNRHVJX9RAstatementtheorem 1.5arxiv
Conditional and final KL-divergencestatementlemma 5.9arxiv
Pinsker's inequality; Lemma 11.6.1statementfact 2.5arxiv
Sequential sampler with $k$ bits of statedefinitiondefinition 4.1arxiv
KL-hardness for conditional samplingdefinitiondefinition 2.22arxiv
Characterization of stateless insertional samplingstatementtheorem 5.6arxiv
Markov chain of sequential sampling statesstatementlemma 4.2arxiv
MY125M11N9statementtheorem 1.4arxiv
Next-block pseudoentropydefinitiondefinition 2.21arxiv
General sampling is stronger than insertional samplingstatementtheorem 7.2arxiv
P5XD719MWRstatementlemma 2.11arxiv
Sorting networksstatementtheorem 2.14arxiv
Batched sampling from a Laplace estimatestatementcorollary 2.16arxiv
KL chain rule for Markov chainsstatementfact 2.8arxiv
Shannon entropydefinitiondefinition 2.1arxiv
PWVXJ4F64Estatementtheorem 1.1arxiv
Binary entropydefinitiondefinition 2.2arxiv
Stateless Insertional Samplerdefinitiondefinition 5.1arxiv
Public-key encryptionstatementcorollary 8.3arxiv
Insertional sampling is stronger than sequential samplingstatementproposition 7.1arxiv
Pseudoentropy, nonuniform settingdefinitiondefinition 2.19arxiv
Conditional pseudoentropy, nonuniform settingdefinitiondefinition 2.20arxiv
Next-block pseudoentropy implies conditional sampling hardnessstatementlemma 3.4arxiv
Separating insertional from ordinary pseudoentropy gapsstatementcorollary 7.4arxiv
Insertional pseudoentropy implies conditional sampling hardnessstatementlemma 5.10arxiv
Insertionally samplable pseudorandom generatorsstatementtheorem 6.1arxiv
KL divergencedefinitiondefinition 2.3arxiv
State elimination for sequential samplers (small KLstatementtheorem 4.3arxiv
Conditional KL bound from joint Laplace estimationstatementlemma 2.12arxiv

Catalogue records