BENCHMARKS DE FLOWING

Dos evaluaciones: calidad de escritura y detección de errores

En 7 disciplinas, 130 artículos distintos y 280 posiciones fijas, Autocompletado con Evidence logró una ventaja global de 31 victorias; Evidence Card detectó el 91,4 % de seis tipos de error explícito.

EVIDENCE CARD · PRUEBA CONTROLADA · REVISIÓN INDEPENDIENTE

Introdujimos errores claros. Evidence Card detectó el 91,4 %.

Detección de errores Red

91,4 %32/35

32/35 confirmados; 33/35 al incluir el caso pendiente.

36 retos Red preparadosRevisión independiente: kimi-k2.6

Precisión de tarjetas Red

97,1 %34/35

34/35 confirmadas · 1 pendiente · 0 rechazada

Alertas Red confirmadas34
46/48 posiciones completadas95,8 %

Controles Green correctos

100,0 %11/11

11/11 controles Green correctos; 0/11 falsas alertas Red.

PRECISIÓN YELLOW · CASOS NATURALES

6/8 confirmadas por revisión independiente

75,0 %

Las tareas y los denominadores son distintos; los resultados se presentan por separado.

MÉTRICAS PUBLICADAS

Detección, precisión de tarjetas, controles y finalización

Cada métrica conserva su propio denominador. Los fallos técnicos se muestran aparte y no cuentan como «sin problema».

46/48 posiciones completadas

Detección de errores Red

91,4 %

32/35 confirmados; 33/35 al incluir el caso pendiente.

Precisión de tarjetas Red

97,1 %

34/35 confirmadas · 1 pendiente · 0 rechazada

Controles Green correctos

100,0 %

11/11 controles Green correctos; 0/11 falsas alertas Red.

Pruebas completadas

95,8 %

46/48 posiciones completadas; 2 fallos técnicos se muestran aparte y no cuentan como “sin problema”.

RESULTADOS POR TIPO DE ERROR

Resultados de los seis tipos de error

Cifras, dirección, causalidad, objeto, ámbito y teoría o definición se muestran por separado.

Oscuro = confirmado; claro = pendiente

Cifras

5/6

Sin fallos técnicos

Dirección o polaridad

6/6

Sin fallos técnicos

Causalidad

4/5

1 fallo técnico mostrado aparte

Objeto incorrecto

5/6

Sin fallos técnicos

Ámbito de aplicación

6/6

Sin fallos técnicos

Teoría o definición

6/6

Sin fallos técnicos

PRECISIÓN YELLOW · CASOS NATURALES

75,0 %+43,8 pp

6/8 confirmadas por revisión independiente

Tarjetas Yellow antes y después del ajuste

La precisión estricta subió del 31,3 % al 75,0 %.

Antes del ajuste se emitieron 16 tarjetas: 5 confirmadas, 1 pendiente y 10 rechazadas. La repetición completa emitió 8: 6 confirmadas, 0 pendientes y 2 rechazadas. Se completaron 50/51 posiciones; 1 fallo técnico se informó aparte.

Repetición completa y revisión independiente finalizadas

48 CASOS PÚBLICOS

Entradas, tarjetas y revisión de cada caso

Cada caso incluye la afirmación evaluada, la fuente fijada, la tarjeta emitida y la revisión independiente. Las omisiones y los fallos técnicos siguen en el registro.

Mostrando 48 casos
REDCifrasDetección confirmada

溶液法原位大面积制备钙钛矿光电薄膜成膜的同步辐射可视化结晶过程研究

该研究报道,单个器件面积约15 cm²的大面积全柔性三维钙钛矿太阳能电池转换效率提升至15.23%。

Abrir caso ↓

Afirmación evaluada

该研究报道,单个器件面积约15 cm²的大面积全柔性三维钙钛矿太阳能电池转换效率提升至15.23%。

Fuente fijada

转换效率提升至5.23%(单个器件面积约15 cm2)

Tarjeta emitida · RED

目标将该研究中约15 cm²全柔性三维钙钛矿太阳能电池的转换效率写为15.23%,而原文对同一器件报告的是5.23%,构成直接数值矛盾。

“基于各层均采用卷对卷 全溶液狭缝涂布方法制备的大面积全柔性三维钙钛矿薄膜太阳能电池转换效率提升至5.23% (单个器件面积 约15 cm2)”— 溶液法原位大面积制备钙钛矿光电薄膜成膜的同步辐射可视化结晶过程研究

Revisión independiente · valid

TARGET QUOTE声称'单个器件面积约15 cm²的大面积全柔性三维钙钛矿太阳能电池转换效率提升至15.23%',SOURCE QUOTE对同一器件(单个器件面积约15 cm²、大面积全柔性三维钙钛矿薄膜太阳能电池)报告转换效率为'5.23%'。两个数值15.23%与5.23%在相同实体(约15 cm²大面积全柔性三维钙钛矿太阳能电池)、相同关系(转换效率)上存在直接数值矛盾,构成直接反驳。

REDCifrasDetección confirmada

Genome-Scale CRISPR-Cas9 Knockout Screening in Human Cells

The GeCKO library targeted 18,080 human genes with 46,751 unique guide sequences.

Abrir caso ↓

Afirmación evaluada

The GeCKO library targeted 18,080 human genes with 46,751 unique guide sequences.

Fuente fijada

targeting 18,080 genes with 64,751 unique guide sequences

Tarjeta emitida · RED

The stated 46,751 unique guide sequences contradicts the source's 64,751 for the same GeCKO library. The source agrees with the stated 18,080 targeted genes.

“We show that lentiviral delivery of a genome-scale CRISPR-Cas9 knockout (GeCKO) library targeting 18,080 genes with 64,751 unique guide sequences enables both negative and positive selection screening in human cells.”— Genome-Scale CRISPR-Cas9 Knockout Screening in Human Cells

Revisión independiente · valid

The TARGET QUOTE states '46,751 unique guide sequences' for the GeCKO library, while the SOURCE QUOTE states '64,751 unique guide sequences' for the same GeCKO library targeting the same 18,080 genes. This is a direct numerical contradiction on the exact same entity (GeCKO library), same relation (number of unique guide sequences), and same scope. The source explicitly confirms the 18,080 genes figure but contradicts the 46,751 guide sequences figure. The proposed 'red' judgment is correct.

REDCifrasDetección confirmada

Empagliflozin, Cardiovascular Outcomes, and Mortality in Type 2 Diabetes

EMPA-REG OUTCOME treated a total of 8,020 patients and followed them for a median of 3.1 years.

Abrir caso ↓

Afirmación evaluada

EMPA-REG OUTCOME treated a total of 8,020 patients and followed them for a median of 3.1 years.

Fuente fijada

A total of 7020 patients were treated (median observation time, 3.1 years)

Tarjeta emitida · RED

The EMPA-REG OUTCOME report states that 7,020 patients were treated, not 8,020.

“A total of 7020 patients were treated (median observation time, 3.1 years).”— Empagliflozin, Cardiovascular Outcomes, and Mortality in Type 2 Diabetes

Revisión independiente · valid

The target quote states '8,020 patients' while the source quote explicitly states '7020 patients were treated.' This is a direct numerical contradiction on the identical claim of patient count in the EMPA-REG OUTCOME trial. The entity (EMPA-REG OUTCOME trial), relation (patient count), and scope (total treated patients) match exactly, with only the number differing. The proposed color red is correct because the source directly contradicts the target proposition on a material point.

REDCifrasDetección confirmada

TabPFN: A Transformer That Solves Small Tabular Classification Problems in a Second

The paper's small-task setting allows up to 10,000 training examples, 100 numerical features, and 10 classes.

Abrir caso ↓

Afirmación evaluada

The paper's small-task setting allows up to 10,000 training examples, 100 numerical features, and 10 classes.

Fuente fijada

≤ 1 000 training examples, ≤ 100 purely numerical features and ≤ 10 classes

Tarjeta emitida · RED

The paper defines its small-task setting as at most 1,000 training examples, not 10,000.

“has thus learned to solve novel small tabular classification tasks (≤ 1 000 training examples,≤ 100 purely numerical features without missing values and≤ 10 classes)”— TabPFN: A Transformer That Solves Small Tabular Classification Problems in a Second

Revisión independiente · valid

The TARGET QUOTE states 'up to 10,000 training examples,' while the SOURCE QUOTE explicitly states '≤ 1 000 training examples.' This is a direct numerical contradiction on the same entity (training examples), same relation (upper bound on small-task setting), same scope (the paper's definition of small-task setting). The source directly contradicts the target proposition that the limit is 10,000 by specifying it is 1,000.

REDCifrasDetección confirmada

The Progress, Challenges, and Perspectives of Directed Greybox Fuzzing

The survey investigates 24 state-of-the-art fuzzers closely related to directed greybox fuzzing.

Abrir caso ↓

Afirmación evaluada

The survey investigates 24 state-of-the-art fuzzers closely related to directed greybox fuzzing.

Fuente fijada

investigation of 42 state-of-the-art fuzzers that are closely related to DGF

Tarjeta emitida · RED

The survey explicitly reports investigating 42 state-of-the-art fuzzers closely related to DGF, not 24.

“Based on the investigation of 42 state-of-the-art fuzzers that are closely related to DGF, we conduct the first in-depth study to summarize the empirical evidence on the research progress of DGF.”— The Progress, Challenges, and Perspectives of Directed Greybox Fuzzing

Revisión independiente · valid

The target quote claims '24 state-of-the-art fuzzers closely related to directed greybox fuzzing.' The source quote explicitly states '42 state-of-the-art fuzzers that are closely related to DGF' (where DGF = directed greybox fuzzing). The entity (fuzzers closely related to directed greybox fuzzing), relation (investigation of), and scope (state-of-the-art) match exactly; the only difference is the number (24 vs. 42), which is a direct numerical contradiction on a material point. The proposed color 'red' is correct.

REDCifrasNo detectado

Evaluating the Impact of Brazil's Bolsa Família: Cash Transfer Programs in Comparative Perspective

The review describes Bolsa Família as benefiting roughly 17 million families.

Abrir caso ↓

Afirmación evaluada

The review describes Bolsa Família as benefiting roughly 17 million families.

Fuente fijada

benefiting roughly 11 million families

No se emitió ninguna tarjeta

REDDirección o polaridadDetección confirmada

Empagliflozin, Cardiovascular Outcomes, and Mortality in Type 2 Diabetes

In EMPA-REG OUTCOME, cardiovascular death was significantly higher with empagliflozin than with placebo.

Abrir caso ↓

Afirmación evaluada

In EMPA-REG OUTCOME, cardiovascular death was significantly higher with empagliflozin than with placebo.

Fuente fijada

significantly lower rates of death from cardiovascular causes (3.7%, vs. 5.9% in the placebo group)

Tarjeta emitida · RED

The claim that cardiovascular death was higher with empagliflozin contradicts the account of the same trial’s cardiovascular results, which explicitly reports a lower rate of death from cardiovascular causes.

“Cardiovascular Outcomes The primary outcome that was reported previ ously showed a significantly lower risk of 3-point MACE in the pooled empagliflozin group than in the placebo group, a difference that was driven by a lower rate of death from cardiovas cular causes.23”— Empagliflozin and Progression of Kidney Disease in Type 2 Diabetes

Revisión independiente · valid

The target quote explicitly claims cardiovascular death was 'significantly higher with empagliflozin than with placebo' in EMPA-REG OUTCOME. The source quote directly contradicts this by stating there was 'a lower rate of death from cardiovascular causes' in the empagliflozin group, which 'drove' the composite benefit. Both quotes concern the same trial (EMPA-REG OUTCOME), the same outcome (cardiovascular death/mortality), the same comparison (empagliflozin vs. placebo), and the same directionality claim—but with opposite directions. The contradiction is material and unambiguous: 'higher' vs. 'lower' for the identical endpoint in the identical trial.

REDDirección o polaridadDetección confirmada

Evaluating the Impact of Brazil's Bolsa Família: Cash Transfer Programs in Comparative Perspective

The review concludes that Bolsa Família increased inequality and extreme poverty.

Abrir caso ↓

Afirmación evaluada

The review concludes that Bolsa Família increased inequality and extreme poverty.

Fuente fijada

helped reduce inequality and extreme poverty

Tarjeta emitida · RED

The review attributes 21 percent of Brazil’s 1995–2004 Gini decline to Bolsa Família, contradicting the claimed increase in inequality.

“They found that the Gini index for Brazil fell by 4.7 percent from 1995 to 2004 and that Bolsa Familia was responsible for 21 percent of that fall (before 1995 there were no CCT programs in Brazil, so there is confidence that 1995 is a baseline year).”— Evaluating the Impact of Brazil's Bolsa Família: Cash Transfer Programs in Comparative Perspective

Revisión independiente · valid

The TARGET QUOTE states 'The review concludes that Bolsa Família increased inequality.' The SOURCE QUOTE from the same review states that Bolsa Família was responsible for '21 percent of that fall' in the Gini index, where 'fall' means a decrease in inequality. A decrease in the Gini index signifies reduced inequality. Therefore, the source directly contradicts the target proposition: the review concludes Bolsa Família decreased inequality (contributed to 21% of the Gini decline), not increased it. The scope matches—both concern the review's conclusion about Bolsa Família's effect on inequality.

1 / 6

Alcance y limitaciones

  • 1Esta prueba controlada mide si el sistema detecta errores explícitos. No estima con qué frecuencia aparecen en artículos reales.
  • 2Algunos artículos fuente también se usaron en pruebas tempranas. Los resultados sirven para regresión y detección controlada, no para estimar prevalencia con una muestra aleatoria.
  • 3La ejecución ajustada emitió solo 8 tarjetas Yellow. Debe mostrarse el denominador 6/8; el 75 % no estima la precisión general.

← Evaluación de autocompletado