Full-paper Knowledge Units · saved output
THE 2-ADIC VALUATION OF STIRLING NUMBERS
Qwen 3.8 27B FP8 · teacher · 1,000-word source chunks · guarded TranslateGemma 4B back-translation
Complete extractionLongest literal run: 15 words
- Source title
- THE 2-ADIC VALUATION OF STIRLING NUMBERS
- Source authors
- Tewodros Amdeberhan; Dante V Manna; Victor H Moll
- KU document title
- THE 2-ADIC VALUATION OF STIRLING NUMBERS
- KU document author field
- Not populated by the original extractor
- Style annotation
- No dedicated document-style field in this schema
- KU sequence
- 4 of 4 expected chunks
- Attribution provenance
- Source registry metadata shown for attribution; not added to the QA context
This condition scores 87.5% (175/200) across all 20 evaluation papers. The example page does not imply that every claim in this individual paper is correct.
Every saved KU, entity, attribute and relationship is shown below, in original chunk order. Navigation follows the source sequence rather than publication dates. Adjacent summaries are reader navigation, not lookahead supplied to extraction. All KUs were generated before the paraphrasing stage. Download complete factual KU JSON.
Original paper: 10.1080/10586458.2008.10129026
KU 1 · source words 1–978
15 entities · 13 relationships · 0 preceding KUs available during extraction
Recorded context summary
The paper analyzes the 2-adic valuations of Stirling numbers of the second kind, building on prior work regarding the sequence b_{l,m} and polylogarithmic series. It defines the Stirling numbers S(n, k) as the count of partitions of an n-element set into k nonempty subsets. A main conjecture is proposed involving an algorithm to describe the function ν_2(S(n, k)), with the specific case k=5 established in the text.
Stirling numbers of the second kind
concept · ID: stirling_numbers_of_the_second_kind- notation
- S(n, k)
- definition
- number of ways to partition a set of n elements into exactly k nonempty subsets
- domain
- n ∈ N and 0 ≤ k ≤ n
Relationships
subject_of_analysis → 2-adic valuation
2-adic valuation
measurement · ID: 2_adic_valuation- notation
- ν_2(m)
- definition
- highest power of 2 that divides m
Relationships
applied_to → Stirling numbers of the second kind
Tewodros Amdeberhan
person · ID: tewodros_amdeberhanNo attributes recorded.
Relationships
author_of → THE 2-ADIC VALUATION OF STIRLING NUMBERS
Dante V Manna
person · ID: dante_v_mannaNo attributes recorded.
Relationships
author_of → THE 2-ADIC VALUATION OF STIRLING NUMBERS
Victor H Moll
person · ID: victor_h_mollNo attributes recorded.
Relationships
author_of → THE 2-ADIC VALUATION OF STIRLING NUMBERS
Legendre's formula
method · ID: legendre_s_formula- formula
- ν_p(m!) = (m - s_p(m)) / (p - 1)
- specific_case_p2
- ν_2(m!) = m - s_2(m)
Relationships
provides_proof_for → Kummer's identity
Kummer's identity
concept · ID: kummer_s_identity- formula
- ν_2(m choose k) = s_2(k) + s_2(m - k) - s_2(m)
Relationships
proved_by → Legendre's formula
sequence b_{l,m}
concept · ID: sequence_b_l_m- definition
- sum from k=l to m of (2^k * binomial(2m-2k, m-k) * binomial(m+k, m) * binomial(k, l))
- origin
- evaluation of definite integral N_{0,4}(a; m)
Relationships
A_{l,m}
concept · ID: a_l_m- definition
- (l! / m!) * 2^{m-l} * b_{l,m}
- valuation_formula
- ν_2(A_{l,m}) = 3l - s_2(m + l) + s_2(m - l)
Relationships
valuation_established_by → J. Little
J. Little
person · ID: j_littleNo attributes recorded.
Relationships
established_valuation_for → A_{l,m}
H. Cohen
person · ID: h_cohenNo attributes recorded.
Relationships
proved_valuation_for → L_1(2^m)proved_valuation_for → L_2(2^m)
L_1(2^m)
measurement · ID: l_1_2_m- valuation_formula
- ν_2(L_1(2^m)) = 2^m + 2m - 4
- condition
- m ≥ 4
Relationships
No relationships recorded.
L_2(2^m)
measurement · ID: l_2_2_m- valuation_formula
- ν_2(L_2(2^m)) = 2^m + m - 1
- condition
- m ≥ 4
Relationships
No relationships recorded.
Main conjecture
concept · ID: main_conjecture- description
- algorithm leading to description of ν_2(S(n, k))
- established_case
- k = 5
Relationships
C_{m,j}
concept · ID: c_m_j- definition
- {2^m * i + j : i ∈ N such that 2^m * i + j ≥ k}
Relationships
No relationships recorded.
Source positions, MinHash and extraction warnings
{
"source": {
"chunk_index": 0,
"start_word": 0,
"end_word": 978,
"word_count": 978,
"sha256": "6b838938ab71be91ad1bea519883c0a2d6d010303c1059bcbbde3463e8a04736",
"sentence_minhash": [
84989070151237717,
1306423125707224,
132898077708225042,
53442166147602330,
148925132169520149,
139865850146007968,
149017575652778032,
34584616103928807,
21579496338166870,
76931577636070685,
78280491562202942,
171596837418463449,
12488864069851856,
58558631451210278,
303264414993015767,
658790404649858
]
},
"extraction_warnings": []
}Previous KU: Start of paper
Next KU: The text defines the partition of integers into classes $C_{m,j}$ based on modulo $2^m$ and introduces the concept of 'm-levels' to analyze the 2-adic valuation of Stirling numbers. It states a main conjecture regarding the stability of class counts and the emergence of constant valuation classes, illustrated with the case $k=11$. Additionally, it reviews known results for specific valuations, including a proven formula for even indices and elementary cases for small $k$.
KU 2 · source words 979–1,977
8 entities · 12 relationships · 1 preceding KUs available during extraction
Recorded context summary
The text defines the partition of integers into classes $C_{m,j}$ based on modulo $2^m$ and introduces the concept of 'm-levels' to analyze the 2-adic valuation of Stirling numbers. It states a main conjecture regarding the stability of class counts and the emergence of constant valuation classes, illustrated with the case $k=11$. Additionally, it reviews known results for specific valuations, including a proven formula for even indices and elementary cases for small $k$.
class C_{m,j}
concept · ID: class_c_m_j- definition
- Set of values $\nu_2(S(2^m i + j, k))$ for $i \in \mathbb{N}$ where $2^m i + j \ge k$
- partition_property
- Forms a partition of $\{n \in \mathbb{N} : n \ge k\}$ into classes modulo $2^m$
Relationships
is_constant_if → valuation consists of a single valuesplits_into → C_{m,ij} and C_{m,ij+2^{m-1}}- condition
- from (m-1)-level class C_{m-1,ij}
m-level
concept · ID: m_level- definition
- Inductively defined set of classes; 1-level consists of even and odd integers $\ge k$
Relationships
defined_by → splitting of (m-1)-level classes
Main Conjecture 1.5
other · ID: main_conjecture_1_5- subject
- 2-adic valuation of Stirling numbers of the second kind $S(n, k)$
- parameter_definition
- $m_0 = m_0(k)$ defined by $2^{m_0-1} < k \le 2^{m_0}$
- model_entity_type
- conjecture
Relationships
predicts_first_constant_level → m_0 - 1- detail
- Each class belongs to the j-level for $1 \le j \le m_0 - 2$
predicts_class_count → 2^{m_0-2}- condition
- for any $m \ge m_0$
- detail
- Each class generates a unique, non-constant class for the (m+1)-level.
S(n, 11)
other · ID: s_n_11- k_value
- 11
- m_0_value
- 4
- m_0_justification
- $2^3 < 11 \le 2^4$
- model_entity_type
- example_case
Relationships
predicted_first_constant_level → 3-levelpredicted_classes_per_level → 4- condition
- for levels $m \ge 4$
2_level_status → non-constant- classes
- C_{2,0}, C_{2,1}, C_{2,2}, C_{2,3}
3_level_valuations → C_{3,3}, C_{3,5} have valuation {0}; C_{3,4}, C_{3,6} have valuation {1}
Lengyel-De Wannemacker result
other · ID: lengyel_de_wannemacker_result- formula
- $\nu_2(S(2^n, k)) = s_2(k) - 1$
- condition
- independent of n
- definition_s_2
- $s_2(k)$ is the sum of binary digits of k
- model_entity_type
- theorem
Relationships
derived_from → Legendre's result- equivalence
- $\nu_2(S(2^n, k)) = \nu_2(2^{k-1}/k!)$
companion_formula → $\nu_2(S(2^n + 1, k + 1)) = s_2(k) - 1$- derivation
- Using recurrence $S(2^n + 1, k + 1) = S(2^n, k) + (k + 1)S(2^n, k + 1)$
De Wannemacker inequality
other · ID: de_wannemacker_inequality- formula
- $\nu_2(S(n, k)) \ge s_2(k) - s_2(n)$
- condition
- $0 \le k \le n$
- model_entity_type
- theorem
Relationships
No relationships recorded.
Stirling numbers of order 1
other · ID: stirling_numbers_of_order_1- formula
- $S(n, 1) = 1$ for all $n \in \mathbb{N}$
- model_entity_type
- mathematical_object
Relationships
No relationships recorded.
Stirling numbers of order 3
other · ID: stirling_numbers_of_order_3- formula
- $S(n, 3) = \frac{1}{2}(3^{n-1} - 2^n + 1)$
- valuation_rule
- $\nu_2(S(n, 3)) = 0$ if n is odd, 1 if n is even
- model_entity_type
- mathematical_object
Relationships
derived_via → recurrence iteration- intermediate_step
- $2^{n-1} = S(n, 3) - \sum_{k=1}^{N-1} 3^k(2^{n-k}-1) - 3^N S(n-N, 3)$
Source positions, MinHash and extraction warnings
{
"source": {
"chunk_index": 1,
"start_word": 978,
"end_word": 1977,
"word_count": 999,
"sha256": "662c917d0dcd810271149f29681749e3f16fc7e8e9ea6278305ee9fb43f8b06b",
"sentence_minhash": [
315931021563575867,
1306423125707224,
23103724853578534,
141557533923110381,
146627570953161212,
428209587723997907,
149017575652778032,
113181735942458166,
228439795799410229,
76931577636070685,
78280491562202942,
139309872259060027,
154708990195788138,
906306679132868,
139254418913316130,
37319280976767778
]
},
"extraction_warnings": [
"normalized unsupported entity type 'conjecture' for 'Main Conjecture 1.5'",
"normalized unsupported entity type 'example_case' for 'S(n, 11)'",
"normalized unsupported entity type 'theorem' for 'Lengyel-De Wannemacker result'",
"normalized unsupported entity type 'theorem' for 'De Wannemacker inequality'",
"normalized unsupported entity type 'mathematical_object' for 'Stirling numbers of order 1'",
"normalized unsupported entity type 'mathematical_object' for 'Stirling numbers of order 3'"
]
}Previous KU: The paper analyzes the 2-adic valuations of Stirling numbers of the second kind, building on prior work regarding the sequence b_{l,m} and polylogarithmic series. It defines the Stirling numbers S(n, k) as the count of partitions of an n-element set into k nonempty subsets. A main conjecture is proposed involving an algorithm to describe the function ν_2(S(n, k)), with the specific case k=5 established in the text.
Next KU: The text establishes the 2-adic valuation of Stirling numbers of the second kind for k=3, k=4, and k=5 using recurrence relations and the ultrametric inequality. It defines 'exceptional indices' for k=5 where specific valuation patterns deviate and proves that the valuation is 1 for n divisible by 8.
KU 3 · source words 1,978–2,971
6 entities · 8 relationships · 2 preceding KUs available during extraction
Recorded context summary
The text establishes the 2-adic valuation of Stirling numbers of the second kind for k=3, k=4, and k=5 using recurrence relations and the ultrametric inequality. It defines 'exceptional indices' for k=5 where specific valuation patterns deviate and proves that the valuation is 1 for n divisible by 8.
Stirling numbers of the second kind S(n, 3)
concept · ID: stirling_numbers_of_the_second_kind_s_n_3- valuation_rule
- 0 if n is odd, 1 if n is even
Relationships
derived_from → recurrence (1.29)proven_via → ultrametric inequality
ultrametric inequality
method · ID: ultrametric_inequality- formula
- ν_2(x1 + x2) ≥ Min {ν_2(x1), ν_2(x2)}
- strictness_condition
- strict unless ν_2(x1) = ν_2(x2)
Relationships
equivalent_to → equation (1.6)
Stirling numbers of the second kind S(n, 4)
concept · ID: stirling_numbers_of_the_second_kind_s_n_4- explicit_formula
- 1/6 (4^(n-1) - 3^n - 3*2^(n+1) - 1)
- valuation_rule
- 1 if n is odd, 0 if n is even
Relationships
derived_from → equation (1.27)
Stirling numbers of the second kind S(n, 5)
concept · ID: stirling_numbers_of_the_second_kind_s_n_5- explicit_formula
- 1/24 (5^(n-1) - 4^n + 2*3^n - 2^(n+1) + 1)
- complexity
- valuation is not easy to compute compared to k=3 and k=4
Relationships
exceptional index
concept · ID: exceptional_index- definition
- index i where q_4i = q_4i+3
- notation
- q_n := ν_2(S(n, 5))
Relationships
first_instance → i = 7- values
- q_28 = 6, q_31 = 7
sequence_examples → {7, 39, 71, 103, ...}
Proposition 3.5
other · ID: proposition_3_5- statement
- ν_2(S(8i, 5)) = 1 for all i ≥ 1
- model_entity_type
- finding
Relationships
proven_using → modulo 32 analysis of 24S(8i, 5)
Source positions, MinHash and extraction warnings
{
"source": {
"chunk_index": 2,
"start_word": 1977,
"end_word": 2971,
"word_count": 994,
"sha256": "a7cfe97cca0170a66906ec3b15975b627583225344bbe3e9c875ea5c29b75502",
"sentence_minhash": [
21703368277289870,
18551641988871988,
69729176314767099,
208823991573236229,
148925132169520149,
428209587723997907,
84576184460003765,
63183425104977485,
113198931733945138,
76931577636070685,
78280491562202942,
221655330234976794,
12488864069851856,
906306679132868,
303264414993015767,
112183611702210919
]
},
"extraction_warnings": [
"normalized unsupported entity type 'finding' for 'Proposition 3.5'"
]
}Previous KU: The text defines the partition of integers into classes $C_{m,j}$ based on modulo $2^m$ and introduces the concept of 'm-levels' to analyze the 2-adic valuation of Stirling numbers. It states a main conjecture regarding the stability of class counts and the emergence of constant valuation classes, illustrated with the case $k=11$. Additionally, it reviews known results for specific valuations, including a proven formula for even indices and elementary cases for small $k$.
Next KU: The text proves that the 2-adic valuation of Stirling numbers S(n, 5) is at least 2 for n of the form 8i+4, and identifies specific congruence classes at the 3-level. It introduces a conjecture by Clarke linking the valuation to the zeros of specific polynomials in the 2-adic integers, providing explicit formulas for k=5, 6, and 7. The main conjecture regarding the stability of class counts is established for the case k=5.
KU 4 · source words 2,972–3,926
9 entities · 10 relationships · 3 preceding KUs available during extraction
Recorded context summary
The text proves that the 2-adic valuation of Stirling numbers S(n, 5) is at least 2 for n of the form 8i+4, and identifies specific congruence classes at the 3-level. It introduces a conjecture by Clarke linking the valuation to the zeros of specific polynomials in the 2-adic integers, providing explicit formulas for k=5, 6, and 7. The main conjecture regarding the stability of class counts is established for the case k=5.
Proposition 3.6
other · ID: proposition_3_6- statement
- ν_2(S(8i + 4, 5)) ≥ 2 for all i ≥ 1
- proof_method
- Analysis of identity modulo 32 using congruences for powers of 2, 3, and 5
- model_entity_type
- theorem
Relationships
applies_to → Stirling numbers of the second kind S(n, 5)
Lengyel's result
other · ID: lengyel_s_result- formula
- ν_2(k!S(n, k)) = k - 1
- conditions
- n = a2^q, a odd, q ≥ k - 2
- model_entity_type
- theorem
Relationships
implies_for_k5 → ν_2(S(n, 5)) = 1 for n = a2^q with q ≥ 3
3-level classes for k=5
concept · ID: 3_level_classes_for_k_5- composition
- {C_{3,4}, C_{3,7}}
Relationships
illustrates → Main Conjecture 1.5
Clarke's conjecture
other · ID: clarke_s_conjecture- statement
- ν_p(k! S(n, k)) = ν_p(T_p(n, k))
- definition_T_p
- T_p(n, k) = sum_{j=0}^k (-1)^{k-j} binom(k, j) j^n, omitting j divisible by p
- model_entity_type
- conjecture
Relationships
Theorem 3.8
other · ID: theorem_3_8- condition
- Assumes Clarke's conjecture holds
- formula_even_n
- ν_2(S(n, 5)) = -1 + ν_2(n - u_0)
- formula_odd_n
- ν_2(S(n, 5)) = -1 + ν_2(n - u_1)
- u_0_definition
- Unique zero of f_{0,5} in 2Z_2
- u_1_definition
- Other zero of f_{0,5} in 1 + 2Z_2
- model_entity_type
- theorem
Relationships
depends_on → Clarke's conjectureuses_function → f_{0,5}(x) = 5 + 10*3^x + 5^x
f_{0,6}
other · ID: f_0_6- expression
- -6 - 20*3^x - 6*5^x
- model_entity_type
- function
Relationships
f_{0,7}
other · ID: f_0_7- expression
- 7 + 35*3^x + 21*5^x + 7^x
- model_entity_type
- function
Relationships
Main Conjecture for k=5
other · ID: main_conjecture_for_k_5- status
- Proven in Section 4
- parameter_m0
- 3
- first_constant_level
- 2
- model_entity_type
- conjecture
Relationships
established_for → Stirling numbers of the second kind S(n, 5)
Empirical approximations for ν_2(S(n, 5))
method · ID: empirical_approximations_for_2_s_n_5- nature
- Derived empirically
- exception_set_I1
- Indices where ν_2(S(m, 5)) ≠ ν_2(f_1(m))
- exception_parity
- Every third element in I_1 is odd
- even_exception_progression
- 256m + 156
Relationships
approximates → ν_2(S(n, 5))
Source positions, MinHash and extraction warnings
{
"source": {
"chunk_index": 3,
"start_word": 2971,
"end_word": 3926,
"word_count": 955,
"sha256": "f93cd1e57deb21c5150220aa5b7975014fa38823c9e286f9d55725340a234e4b",
"sentence_minhash": [
21703368277289870,
1306423125707224,
229329608839330056,
53442166147602330,
148925132169520149,
62667647458868521,
12025634198112669,
96292850158067112,
121868570449012213,
92715222163426349,
78280491562202942,
139309872259060027,
8649321992247620,
58558631451210278,
194885768149381016,
69242845566341429
]
},
"extraction_warnings": [
"normalized unsupported entity type 'theorem' for 'Proposition 3.6'",
"normalized unsupported entity type 'theorem' for \"Lengyel's result\"",
"normalized unsupported entity type 'conjecture' for \"Clarke's conjecture\"",
"normalized unsupported entity type 'theorem' for 'Theorem 3.8'",
"normalized unsupported entity type 'function' for 'f_{0,6}'",
"normalized unsupported entity type 'function' for 'f_{0,7}'",
"normalized unsupported entity type 'conjecture' for 'Main Conjecture for k=5'"
]
}Previous KU: The text establishes the 2-adic valuation of Stirling numbers of the second kind for k=3, k=4, and k=5 using recurrence relations and the ultrametric inequality. It defines 'exceptional indices' for k=5 where specific valuation patterns deviate and proves that the valuation is 1 for n divisible by 8.
Next KU: End of paper