Memory Research Articles

OLM-SLAM: Online lifelong memory system for simultaneous localization and mapping

Abstract Simultaneous Localization and Mapping is a fundamental task for robots in unknown environments. However, the poor generalization ability of learning-based algorithms in unknown environments hinders their widespread adoption. Additionally, artificial neural networks are subject to catastrophic forgetting. We propose a lifelong SLAM framework called OLM-SLAM that effectively solves the neural network catastrophic forgetting problem. To ensure the generalization of the neural network, this paper proposes a method for the sensitivity analysis of the network weight parameters. Meanwhile, inspired by human memory storage mechanisms, we design a dual memory storages mechanism that retains dynamic memory and static memory. A novel memory filtering mechanism is proposed to maximize image diversity within a fixed-size memory storage area addressing the problem of limited storage capacity of embedded devices in real-world situations. We have extensively evaluated the model on a variety of real-world datasets. Compared with CL-SLAM, the overall translation error of the test sequence is improved by 44.9%. The translation and rotation errors of Retention Ability were improved by 111.6% and 66.7%, respectively. The results demonstrate that OLM-SLAM can outperform previous methods of the same type, and OLM-SLAM has high retention ability when facing different sequences of the same type of environment.

Proof of life: Human remains and memory in the Philippine Drug War

This paper examines the aftermath of the Philippine government’s anti-drug campaign under President Rodrigo Duterte, focusing on the impact on families of victims of extrajudicial killings (EJKs). Drawing on fieldwork conducted in Manila, the research explores how these families navigate loss and attempt to reclaim dignity and humanity for their loved ones. The study reveals the formation of a community of "necro-activists" comprising journalists, religious figures, and medical professionals who support victims’ families in seeking justice and remembrance. Through interviews with photojournalists and victims’ families. This paper highlights the transformative role of human remains in asserting the victims’ agency beyond death. The research shows how forensic processes and religious rituals contribute to the reclamation of individual identities and challenge the state’s dehumanizing narratives. By examining the evolving meanings of EJK victims’ remains, this study sheds light on the broader socio-political implications of state violence and memory-making in the Philippines. The analysis situates this research within a global context of human rights and memory studies, drawing parallels with other regions affected by state-sponsored violence. Ultimately, this paper argues that the remains of EJK victims serve as powerful symbols that challenge state impunity and embody the resilience of communities seeking accountability and dignity amid pervasive violence.

Two-dimensional neural geometry underpins hierarchical organization of sequence in human working memory.

Working memory (WM) is constructive in nature. Instead of passively retaining information, WM reorganizes complex sequences into hierarchically embedded chunks to overcome capacity limits and facilitate flexible behaviour. Here, to investigate the neural mechanisms underlying hierarchical reorganization in WM, we performed two electroencephalography and one magnetoencephalography experiments, wherein humans retain in WM a temporal sequence of items, that is, syllables, which are organized into chunks, that is, multisyllabic words. We demonstrate that the one-dimensional sequence is represented by two-dimensional neural representational geometry in WM arising from left prefrontal and temporoparietal regions, with separate dimensions encoding item position within a chunk and chunk position in the sequence. Critically, this two-dimensional geometry is observed consistently in different experimental settings, even during tasks not encouraging hierarchical reorganization in WM and correlates with WM behaviour. Overall, these findings strongly support that complex sequences are reorganized into factorized multidimensional neural representational geometry in WM, which also speaks to general structure-based organizational principles given WM's involvement in many cognitive functions.

Convolutional neural networks uncover the dynamics of human visual memory representations over time.

The ability to accurately retrieve visual details of past events is a fundamental cognitive function relevant for daily life. While a visual stimulus contains an abundance of information, only some of it is later encoded into long-term memory representations. However, an ongoing challenge has been to isolate memory representations that integrate various visual features and uncover their dynamics over time. To address this question, we leveraged a novel combination of empirical and computational frameworks based on the hierarchal structure of convolutional neural networks and their correspondence to human visual processing. This enabled to reveal the contribution of different levels of visual representations to memory strength and their dynamics over time. Visual memory strength was measured with distractors selected based on their shared similarity to the target memory along low or high layers of the convolutional neural network hierarchy. The results show that visual working memory relies similarly on low and high-level visual representations. However, already after a few minutes and on to the next day, visual memory relies more strongly on high-level visual representations. These findings suggest that visual representations transform from a distributed to a stronger high-level conceptual representation, providing novel insights into the dynamics of visual memory over time.

A Deep Learning Model to Predict Breast Implant Texture Types Using Ultrasonography Images: Feasibility Development Study.

Breast implants, including textured variants, have been widely used in aesthetic and reconstructive mammoplasty. However, the textured type, which is one of the shell texture types of breast implants, has been identified as a possible etiologic factor for lymphoma, specifically breast implant-associated anaplastic large cell lymphoma (BIA-ALCL). Identifying the shell texture type of the implant is critical to diagnosing BIA-ALCL. However, distinguishing the shell texture type can be difficult due to the loss of human memory and medical history. An alternative approach is to use ultrasonography, but this method also has limitations in quantitative assessment. This study aims to determine the feasibility of using a deep learning model to classify the shell texture type of breast implants and make robust predictions from ultrasonography images from heterogeneous sources. A total of 19,502 breast implant images were retrospectively collected from heterogeneous sources, including images captured from both Canon and GE devices, images of ruptured implants, and images without implants, as well as publicly available images. The Canon images were trained using ResNet-50. The model's performance on the Canon dataset was evaluated using stratified 5-fold cross-validation. Additionally, external validation was conducted using the GE and publicly available datasets. The area under the receiver operating characteristic curve (AUROC) and the area under the precision-recall curve (PRAUC) were calculated based on the contribution of the pixels with Gradient-weighted Class Activation Mapping (Grad-CAM). To identify the significant pixels for classification, we masked the pixels that contributed less than 10%, up to a maximum of 100%. To assess the model's robustness to uncertainty, Shannon entropy was calculated for 4 image groups: Canon, GE, ruptured implants, and without implants. The deep learning model achieved an average AUROC of 0.98 and a PRAUC of 0.88 in the Canon dataset. The model achieved an AUROC of 0.985 and a PRAUC of 0.748 for images captured with GE devices. Additionally, the model predicted an AUROC of 0.909 and a PRAUC of 0.958 for the publicly available dataset. This model maintained the PRAUC values for quantitative validation when masking up to 90% of the least-contributing pixels and the remnant pixels in breast shell layers. Furthermore, the prediction uncertainty increased in the following order: Canon (0.066), GE (0072), ruptured implants (0.371), and no implants (0.777). We have demonstrated the feasibility of using deep learning to predict the shell texture type of breast implants. This approach quantifies the shell texture types of breast implants, supporting the first step in the diagnosis of BIA-ALCL.

Modeling collaborative memory with SAM.

While humans often encode and retrieve memories in groups, the bulk of our knowledge of human memory comes from paradigms with individuals in isolation. The primary phenomenon of interest within the relatively new field of collaborative memory is collaborative inhibition: the tendency for collaborative groups to underperform in free recall tasks compared with noncollaborative groups of the same size. This effect has been found in a variety of materials and group compositions. However, most research in this field is led by verbal conceptual theories without guidance from formal computational models. We present a framework to scale the Search of Associative Memory model (SAM) to collaborative free recall paradigms with multiple models working together. Multiple SAM models recalling together naturally produce collaborative inhibition when the group members use recalls by the group as cues to retrieve from memory, strongly supporting the "retrieval disruption" hypothesis. This work shows that SAM can act as a unified theory to explain both individual and collaborative memory effects and offers a framework for future predictions of scaling to increased group sizes, shared knowledge, and factors facilitating the spread of false memories in groups.

Repulsion of CA3 / dentate gyrus representations is driven by distinct internal beliefs in the face of ambiguous sensory input.

Recent human neuroimaging studies of episodic memory have revealed a counterintuitive phenomenon in the hippocampus: when events are highly similar, corresponding hippocampal activity patterns are sometimes less correlated than activity patterns associated with unrelated events. This phenomenon- repulsion- is not accounted for by most theories of the hippocampus, and the conditions that trigger repulsion remain poorly understood. Here, we used a spatial route-learning task and high-resolution fMRI in humans to test whether hippocampal repulsion is fundamentally driven by internal beliefs about the environment. By precisely measuring participants' internal beliefs and actively manipulating them, we show that repulsion selectively occurred in hippocampal subfields CA3 and dentate gyrus when visual input was ambiguous-or even identical -but internal beliefs were distinct. These findings firmly establish conditions that elicit repulsion and have broad relevance to theories of hippocampal function and to the fields of human episodic memory and rodent spatial navigation.

Multiplexed Antibody Sequencing and Profiling of the Human Hemagglutinin-specific Memory B Cell Response following Influenza Vaccination.

Influenza virus is a highly contagious respiratory pathogen causing between 9.4 and 41 million infections per year in the United States in the last decade. Annual vaccination is recommended by the World Health Organization, with the goal to reduce influenza severity and transmission. Ag-specific single B cell sequencing methodologies have opened up new avenues into the dissection of the Ab response to influenza virus. The improvement of these methodologies is pivotal to reduce the associated costs and optimize the operational workflow and throughput, especially in the context of multiple samples. In this study, PBMCs and serum samples were collected longitudinally from eight influenza vaccinees either vaccinated yearly for four consecutive influenza seasons or once for one season. Following the serological and B cell profiling of their polyclonal Ab response to a panel of historical, recent, and next-generation influenza vaccine hemagglutinin (HA) and virus strains, a single multiplexed Ag-specific single B cell sequencing run allowed to capture HA-specific memory B cells that were analyzed for preferential Ig H chain/L chain pairing, isotype/subclass usage, and the presence of public BCR clonotypes across participants. Binding and functional profiles of representative private and public clonotypes confirmed their HA specificity, and their overall binding and functional activity were consistent with those observed at the polyclonal level. Collectively, this high-resolution and multiplexed Ab repertoire analysis demonstrated the validity of this optimized methodology in capturing Ag-specific BCR clonotypes, even in the context of a rare B cell population, such as in the case of the peripheral Ag-specific memory B cells.

Executive control can query hidden human memories.

When we try to retrieve a representation from visual long-term memory there is a chance that we will fail to recall seeing it even though the memory is stored in our brain. Here we show that although mechanisms of explicit memory retrieval are sometimes unable to retrieve stored memories, that mechanisms of executive control can quickly query memory and determine if a representation is stored therein. Our findings suggest that the representations stored in human memory that cannot be accessed explicitly at that moment are nonetheless directly accessible by the brain's higher level control mechanisms.

Exploring the links among brain iron accumulation, cognitive performance, and dietary intake in older adults: A longitudinal MRI study

This study evaluated longitudinal brain iron accumulation in older adults, its association with cognition, and the role of specific nutrients in mitigating iron accumulation. MRI-based, quantitative susceptibility mapping estimates of brain iron concentration were acquired from seventy-two healthy older adults (47 women, ages 60–86) at a baseline timepoint (TP1) and a follow-up timepoint (TP2) 2.5–3.0 years later. Dietary intake was evaluated at baseline using a validated questionnaire. Cognitive performance was assessed at TP2 using the uniform data set (Version 3) neuropsychological tests of episodic memory (MEM) and executive function (EF). Voxel-wise, linear mixed-effects models, adjusted for longitudinal gray matter volume alterations, age, and several non-dietary lifestyle factors revealed brain iron accumulation in multiple subcortical and cortical brain regions, which was negatively associated with both MEM and EF performance at T2. However, consumption of specific dietary nutrients at TP1 was associated with reduced brain iron accumulation. Our study provides a map of brain regions showing iron accumulation in older adults over a short 2.5-year follow-up and indicates that certain dietary nutrients may slow brain iron accumulation.

Distinct functions for beta and alpha bursts in gating of human working memory

Multiple neural mechanisms underlying gating to working memory have been proposed with divergent results obtained in human and animal studies. Previous findings from non-human primates suggest prefrontal beta frequency bursts as a correlate of transient inhibition during selective encoding. Human studies instead suggest a similar role for sensory alpha power fluctuations. To cast light on these discrepancies we employed a sequential working memory task with distractors for human participants. In particular, we examined their whole-brain electrophysiological activity in both alpha and beta bands with the same single-trial burst analysis earlier performed on non-human primates. Our results reconcile earlier findings by demonstrating that both alpha and beta bursts in humans correlate with the filtering and control of memory items, but with region and task-specific differences between the two rhythms. Occipital beta burst patterns were selectively modulated during the transition from sensory processing to memory retention whereas prefrontal and parietal beta bursts tracked sequence order and were proactively upregulated prior to upcoming target encoding. Occipital alpha bursts instead increased during the actual presentation of unwanted sensory stimuli. Source reconstruction additionally suggested the involvement of striatal and thalamic alpha and beta. Thus, specific whole-brain burst patterns correlate with different aspects of working memory control.

A third kind of episodic memory: Context familiarity is distinct from item familiarity and recollection.

Memory for past events is widely believed to operate through two different processes: one called 'recollection' when retrieving confident, specific details of a memory, and another called 'familiarity' when only having an unsure but conscious awareness that an item was experienced before. When people successfully retrieve details such as the source or context of a prior event, it has been assumed to reflect recollection. We demonstrate that familiarity of context is functionally distinct from familiarity of items and recollection and offer a new, tri-component model of memory. The three memory responses were differentiated across multiple behavioral and brain wave measures. What has traditionally been thought to be two kinds of memory processes are actually three, becoming evident when using sensitive enough multi-measures. Results are independently replicated across studies from different labs. These data reveal that context familiarity is a third process of human episodic memory.

Eyewitness Testimony – How Reliable is It?

This paper examines the intricacies associated with eyewitness testimony in legal settings, emphasizing its crucial role as evidence while also addressing its fundamental limitations. An eyewitness is described as an individual who witnesses an event firsthand, and their description can greatly influence the court's decisions. Nevertheless, factors such as stress, racial bias, focus on firearms, and memory lapses can compromise the accuracy of these accounts. The article analyses the legal framework established by the Bharatiya Sakshya Adhiniyam, 2023, which pertains to the competency of witnesses and the reliability of their testimonies. An analysis of the prominent wrongful conviction case of Ronald Cotton, highlights the fallibility of human memory. Furthermore, a review of pertinent psychological studies indicates that eyewitness memory is reconstructive and susceptible to inaccuracies. The necessity for corroborative evidence, expert opinions, and enhanced identification processes is highlighted as a means to improve the reliability of eyewitness testimonies and reduce wrongful convictions, thereby upholding the integrity of the trial process. In summary, this study advocates for a thorough reassessment of eyewitness testimony within a wider legal context to ensure justice.

Amorphous In–Al–Sn–O Thin Film Transistors and Their Application in Optoelectronic Artificial Synapses

AbstractIn‐Al‐Sn‐O (IATO) is a very promising novel amorphous oxide as the active layer of thin film transistors (TFTs). Herein, IATO TFTs are first fabricated with the effects of annealing on IATO films and TFTs being studied. The IATO films possessed amorphous structure, flat surface morphology, high visible light transmittance, and wide optical bandgap ≈4.20 eV before and after annealing even at 400 °C. The minimal surface roughness and internal defects are obtained for the 300 °C annealed IATO film. Correspondingly, the 300 °C annealed TFTs demonstrated the best overall performance including high saturation mobility (8.55 ± 0.62 cm2 V−1 s−1), low subthreshold swing (0.40 ± 0.07 V dec−1), ideal on/off current ratio (1.25 ± 0.09 × 108), and negligible hysteresis (0.23 ± 0.03 V) values. The 300 °C annealed TFTs are then applied in optoelectronic artificial synapses and exhibit typical synaptic properties, including excitatory postsynaptic current, paired‐pulse facilitation, and short‐term plasticity to long‐term plasticity conversion in response to light stimulation. The international Morse code and repetitive learning‐forgetting behavior of the human brain are also successfully simulated. In particular, an emotion‐memory efficiency model is proposed and the emotion effect on human memory efficiency is successfully imitated via the regulation of gate voltage.

Remembering as an operant: Effects of instructional control and reinforcement on remembering behavior.

With the general aim of providing more evidence for considering certain behaviors involved in the act of remembering as operant, two experiments were carried out to verify its sensitivity to differential reinforcement, and to some of the variables upon which it depends. In the first experiment, three children participated, and two variables were manipulated in a "free recall" task: the accuracy of the instructions and the magnitude of the reinforcer applied to the emission of the target words. In the second experiment, 60 was changed to one of comparison between groups. In this case, the response-reinforcer interval (immediate vs. delayed) was manipulated using a "recognition" task. In both experiments, a greater number of remembered items were found in the presence of reinforcement compared to its absence, as well as with a greater magnitude of the reinforcer and with a shorter delay. These results are discussed considering the peculiarities of verbal behavior in humans and provide evidence that operant processes have an important role in traditional human memory tasks.

Follicular lymphoma comprises germinal center–like and memory-like molecular subtypes with prognostic significance

AbstractA robust prognostic and biological classification for newly diagnosed follicular lymphoma (FL) using molecular profiling remains challenging. FL tumors from patients treated in the RELEVANCE trial with rituximab-chemotherapy (R-chemo) or rituximab-lenalidomide (R2) were analyzed using RNA sequencing, DNA sequencing, immunohistochemistry (IHC), and/or fluorescence in situ hybridization. Unsupervised gene clustering identified 2 gene expression signatures (GSs) enriched in normal memory (MEM) B cells and germinal center (GC) B-cell signals, respectively. These 2 GSs were combined into a 20-gene predictor (FL20) to classify patients into MEM-like (n = 160) or GC-like (n = 164) subtypes, which also displayed different mutational profiles. In the R-chemo arm, patients with MEM-like FL had significantly shorter progression-free survival (PFS) than patients with GC-like FL (hazard ratio [HR], 2.13; P = .0023), and this prognostic correlation remained significant in a multivariable model that included the FL-International Prognostic Index (P = .005). In the R2 arm, both subtypes had comparable PFS, demonstrating that R2 has a benefit over R-chemo for patients with MEM-like FL (HR, 0.54; P = .011). The prognostic value of FL20 was validated in an independent FL cohort with R-chemo treatment (GSE119214 [n = 137]). An IHC algorithm (FLCM) that used FOXP1, LMO2, CD22, and MUM1 antibodies was developed with significant prognostic correlation with FL20 in a training set of patients from the RELEVANCE trial (n = 264), which was then validated in a different set of patients (n = 116). These data indicate that FL tumors can be classified into MEM-like and GC-like subtypes that are biologically distinct and clinically different in their risk profile. The FLCM assay can be used in routine clinical practice to identify patients with MEM-like FL who might benefit from therapies other than R-chemo, such as the R2 combination. This trial was registered at www.clinicaltrials.gov as #NCT01476787 and #NCT01650701.

CD27 is not an ideal marker for human memory B cells and can be modulated by IL-21 upon stimulated by Anti-CD40

B cells play a key role in humoral immune responses by producing antibodies. Although there are numerous research on memory B cells definition markers and cytokines on B cell development, different studies have yielded contradictory conclusions due to species studied, the different cells and stimulating agents used. In the current study, we conducted a detailed characterization of B cells in human CBMCs, PBMCs and tonsil, including expression of Igs, activation and memory markers. Furthermore, we found that considerable amounts of IgA and IgG were expressed by CD27− B cells. These “Atypical” memory B cells corresponded to approximately 50% of IgG+ and IgA+B cells in blood, this proportion even reached 90% in tonsil. In addition, we investigated the effect of IL-21 and TGF-β1 on the membrane-bound form and secreted form of Igs using PBMCs and purified blood B cells. There were actual differences between the effect of cytokines on Igs secretion and surface expression. Our study will be helpful to advance the knowledge and understanding of humoral memory.

The β2-adrenergic biased agonist nebivolol inhibits the development of Th17 and the response of memory Th17 cells in an NF-κB-dependent manner.

Adrenergic receptors regulate metabolic, cardiovascular, and immunological functions in response to the sympathetic nervous system. The effect of β2-adrenergic receptor (AR) as a high expression receptor on different subpopulations of T cells is complex and varies depending on the type of ligand and context. While traditional β2-AR agonists generally suppress T cells, they potentially enhance IL-17A production by Th17 cells. The effects of pharmacological drugs that count as biased agonists of AR like nebivolol are not completely understood. We investigated the impact of nebivolol on human memory CD4+ T (Th1, Th2, Th17) cells and polarized naive Th17 cells, highlighting its potential for IL-17A suppression via a non-canonical β2-AR cell signaling pathway. The effects of nebivolol were tested on healthy human peripheral blood mononuclear cells, purified memory Th cells, and polarized naive Th17 cells activated with anti-CD3/anti-CD28/anti-CD2 ImmunoCult reagent. IFN-γ, IL-4, and IL-17A, which are primarily derived from Th1, Th2, and Th17 cells, respectively, were quantified by ELISA and flow cytometry. IL-10 was measured by ELISA. Gene expression of RORC, ADRB1, ADRB2, and ADRB3 was evaluated by qPCR. The ADRB2 gene was knocked out in memory Th cells using CRISPR/Cas9. Protein expression of phosphorylated serine133-CREB and phosphorylated NF-κB p65 was assessed by Western blot. Proliferation was assessed by fluorescent dye loading and flow cytometry. Nebivolol treatment decreased IL-17A and IFN-γ secretion by activated memory Th cells and elevated IL-4 levels. Nebivolol reduced the proportion of IL-17A+ Th cells and downregulated RORC expression. Unlike the β2-AR agonist terbutaline, nebivolol inhibited the shift of naive CD4+ T cells toward the Th17 phenotype. IL-10 and the proliferation index remained unchanged. Nebivolol-treated β2-knockout memory Th cells showed significant inhibition of β2-AR-mediated signaling, evidenced by the absence of IL-17A suppression compared to controls. Phosphorylation of the NF-κB p65 subunit was inhibited by nebivolol, but CREB phosphorylation was not changed, suggesting a selective transcriptional control. The findings demonstrate that nebivolol acts through a β2-AR-mediated signaling pathway, as a distinctive anti-inflammatory agent capable of selectively shifting Th17 cells and suppressing the phosphorylation of NF-κB. This highlights nebivolol's potential for therapeutic interventions in chronic autoimmune conditions with elevated IL-17A levels.

Don't Throw Away Your Notebook: Effects of Task Difficulty and Presentation Medium on Memory Performance.

People are increasingly reliant on various electrical devices for learning and memory, yet the implications and consequences of this dependence remain poorly understood. The present study aimed to investigate how learning through electrical media impacts recall under varying task difficulties. During this study, participants encoded information related to daily life situations (low difficulty), academic conceptual knowledge (middle difficulty), or associative word pairs (high difficulty), presented on smartphones, computers, or paper. At test, they recalled the omitted content based on the provided cue information. A significant screen-inferiority effect was observed for both computers and smartphones. However, the impairment related to computers disappeared in the retrieval of daily life situations and academic conceptual knowledge, whereas the impairment associated with smartphones was consistently present across all tasks. These results suggest that memory performance is modulated by the interaction between the presentation medium and the specific demands of the task, highlighting a more pronounced screen-inferiority memory effect when the media restrict the depth of processing or when the memory tasks pose greater external challenges. A deeper understanding of these factors can guide the optimization of electrical devices to enhance human memory abilities and functions.

Vladimír Vernadskij o nemecko-rakúskej okupácii Ukrajiny v roku 1918

The article deals with the diaries as an important narrative source for historians in their study of the past. The diaries are a very peculiar type of source that is different from memoirs. Usually, the author writes his memoirs about the events that he lived through many years ago or he witnessed in the past. It is naturally that with the flow of time he may forget about something, miss something or recall something incorrectly; moreover, as it is known, in memoirs events of the past are reflected with a hindsight and that is why they do not always accurately reflect the feelings of the author at the time when these events took place. People’s attitudes towards the events can also change as the years go by. Moreover, human memory is very much selective. Unlike memoirs, the diaries allow us to understand the attitude of a particular human being to the surrounding world and contemporary events when he or she wrote about them. This has enormous value for the researcher. The diaries, as a type of historical document, are important precisely because they provide us not only with object of a historical process, in other words, what is being described, but also with insights about their authors who are present not as abstract depersonalized subjects of this process, but as living human beings with their merits and shortcomings, with their convictions and moods, their dreams and concerns. Moreover, the diaries provide unique material for the study of social psychology of people in concrete historical period. The author of diaries, which became the object of our attention, was Vladimir Ivanovich Vernadskyi (1863-1945), an outstanding Russian 20th century naturalist, geologist, biogeochemist, thinker and winner of the Stalin Prize, academician, creator of the doctrine of the biosphere, its evolution and transformation into the noosphere, who was one of the founders of the Ukrainian Academy of Sciences and its first President, as well as the founder of the National Library of Ukraine. Their main value of his diaries, in our opinion, is that they help to feel the atmosphere of the dramatic events of October Revolution and Civil War in Russia and Ukraine, allows us to restore the spirit of the revolutionary stormy years in all its tragedy and contradictions. As for the content of the great scientist’s diaries, a significant part of them consists of the facts of his personal life., but also a lot of space is devoted to his impressions and reflections on political events, in particular about the German and the Austro-Hungarian occupation of Ukraine in 1918. In his diaries, a negative attitude towards the Bolsheviks and socialist ideology, non-acceptance of manifestations of Ukrainian nationalism, and condemnation of the talentless policy of the Central Rada can be clearly traced.