Automatic narrative generation is garnering significant interest in artificial intelligence. Research has explored methods such as repurposing existing literature and the agent-based simulation. The ...rise of large language models (LLMs) has notably advanced this field. In this study, we introduce an LLM-based narrative generation technique via the agent-based simulation (ABS). Within the ABS framework, we employ LLM to create agents with distinct personas including names, ages, and personalities. We demonstrate a method to craft narratives by interacting with these agents. Focusing on a classic "dragon-slaying" scenario typical in RPGs, we generated agents representing characters such as a brave hero, a warrior, a wizard, and a dragon, and weaved a narrative around their roles. By applying the rich expressive power of LLMs to story generation with the ABS, a variety of creative works can be generated. From the experimental results, it is confirmed that the story may not reach a peaceful ending or describe the battle scene in detail due to the influence of reinforcement learning based on human feedback of LLMs. However, the expected results were obtained in terms of the conversation before and after the battle and the agent's autonomous consideration of motives. As a future prospect, we expect that there will be an increase in the number of requests to specifically generate battle scenes when generating stories. For this reason, introduction of another LLM model specifically trained to describe combat scenes, or an approach that presents choices such as "fight," "escape," "use item," "use magic," etc. similar to a typical RPG game screen, and progresses the story according to the choices, may also be considered.
We attempted to predict successful myocardial reperfusion and salvage in acute myocardial infarction (AMI) by using measurements of plasma superoxide dismutase (SOD), a plasma free-radical scavenger, ...activity. Forty-nine patients with AMI were studied within 6 hours of symptoms onset. In group 1 (
n = 26), primary percutaneous transluminal coronary angioplasty (PTCA) was undertaken, and plasma SOD activity was measured for 8 hours by the nitrite method. Left ventricular (LV) angiography was assessed before and 3 months after PTCA by computer LV contraction analysis. In group 2 (
n = 23), TPA was infused intravenously over a 60-minute period, and plasma SOD activity was measured before and immediately after TPA infusion. In group 1, occluded coronary arteries were successfully dilated in 24 of 26 patients, and plasma SOD activity increased from 3.20 ± 0.17 U/ml to 4.66 ± 0.29 U/ml at 1 hour after PTCA (
p < 0.001), returning to the basal level by 8 hours after PTCA. Plasma SOD activity did not significantly change in patients with unsuccessful PTCA or those with the no-reflow phenomenon. The maximal increase in plasma SOD activity was significantly correlated with the grade of improvement in LV contraction (
r = 0.852,
p < 0.001). In group 2, the sensitivity and specificity of predicting coronary recanalization was 86% and 89%, respectively. In conclusion, myocardial reperfusion and salvage in AMI can be predicted by changes in plasma SOD activity.
T cell lymphoma carrying Epstein Barr virus (EBV+TL) is very rare among Western countries while it is much more common among Japanese. Here we report an EBV+TL which has been maintained for years by ...the use of mice with severe combined immune deficiency (SCID) mice. Lymphoma was obtained from a 55‐year‐old male suffering from oculomotor nerve palsy and lymphadenopathy. A small piece of biopsied tumor was transplanted into SCID mice and the lymphoma has been maintained for over 3 years with passages every 2–3 weeks. The maintained lymphoma, termed as TMS24, and the original lymphoma cells showed identical phenotype and genotype, including diffuse medium‐sized cell morphology lacking granules, suppressor/cytotoxic immunophenotype and identical T cell receptor β‐chain gene rearrangement mode. Further, both were shown to carry an identical EBV clone in terms of the number of terminal repeats and the latency II‐type restricted gene expression profile. Cytogenetically, TMS24 retained two characteristic chromosomal translocations of t(l;18)(q32;q21) and t(6;12)(p21;q24). Since only one cell line with such characters has been reported previously, TMS24 should be useful for detailed analysis of EBV+TL.
Ninety-one brain tissue sections taken at autopsy from 33 elderly patients (63-100 years old) without progressive multifocal leukoencephalopathy were examined for the presence of JC virus DNA by the ...polymerase chain reaction and Southern hybridization analysis after DNA extraction. JC virus DNA was detected in 15 sections from 10 patients. These results suggest that JC virus is frequently present in the brains of aged patients.
19 zu 19: Eine konvergente Totalsynthese von 19‐Hydroxysarmentogenin geht von drei einfachen Fragmenten aus. Das gewünschte Produkt wurde in 19 Stufen ausgehend vom AB‐Ring unter Aufbau von sechs ...stereogenen Zentren und der Bildung dreier C‐C‐Bindungen synthetisiert.
Polyoma virus nephropathy after transplantation is believed to be primarily due to the BK virus. We hypothesized that some cases may be associated with the JC polyoma virus (JCV), which is also known ...to be latent in the kidney.
We sought polymerase chain reaction evidence of JCV infection in needle biopsy specimens with and without viral nephropathy. Cases positive by polymerase chain reaction were studied by immunohistochemistry for VP-1 antigen expression.
JCV DNA was found in 7 (36.8%) of 19 allograft kidney biopsy specimens with viral nephropathy and 0 (0%) of 19 native or allograft biopsy specimens without viral nephropathy. Immunohistochemistry localized JCV to the nuclei of tubular epithelial cells in one case.
JCV is detectable in a subset of renal allograft kidneys with polyoma virus nephropathy. The tubular epithelium is identified as a site capable of supporting JCV viral capsid protein VP-1 expression, and hence viral replication.
The protective effects of preinfarction angina were evaluated in acute myocardial infarction (AMI) treated by primary percutaneous transluminal coronary angioplasty (PTCA) and stenting. We studied ...613 patients with AMI. Group 1 (n = 306) was treated by conventional medical therapies and coronary thrombolysis and group 2 (n = 307) was treated by primary PTCA supported by stenting. Each group was subdivided into those with and without preinfarction angina within 24 hours before the onset of AMI. There was no significant difference in clinical characteristics between the subgroups of groups 1 and 2. In group 1, there were differences between patients with preinfarction angina (n = 84) and those without (n = 222) in in-hospital mortality (11% vs 18%), pump failure (Killip classes 3 and 4) (11% vs 21%, p <0.05), left ventricular ejection fraction at discharge (52 ± 13% vs 48 ± 14%, p <0.05), and peak creatine kinase (2,106 ± 1,637 vs 2,764 ± 2,154 U/L, p <0.02). In group 2, however, there was no significant difference between those with preinfarction angina (n = 82) and those without (n = 225) in mortality (6% vs 6%), pump failure (12% vs 12%), left ventricular ejection fraction (50 ± 13% vs 50 ± 13%) and peak creatine kinase (3,285 ± 2,306 vs 3,291 ± 2,262 U/L). Multivariate analysis indicated that preinfarction angina was an independent determinant of in-hospital death and pump failure in group 1, but not in group 2. We conclude that the protective effects of preinfarction angina in AMI are not evident in those treated by primary PTCA and stenting, possibly because of the overwhelming protective effects of complete coronary revascularization provided by primary PTCA and stenting.
PDF
