CoolFace
Datasetpublic

AmazonScience/document-haystack

Document Haystack Dataset This repository contains the dataset for the paper “Document Haystack: A Long Context Multimodal Image/Document Understanding Vision LLM Benchmark”. 📑 Abstract Paper The proliferation of multimodal Large Language Models has significantly advanced the ability to analyze and understand complex data inputs from different modalities. However, the processing of long documents remains under-explored, largely due to a lack of suitable… See the full description on the dataset page: https://huggingface.co/datasets/AmazonScience/document-haystack.

sourceHugging Faceupdated 1y agoView on Hugging Face
20likes90kdownloads
AIG_75Pages_TextNeedles_page_43.txt57 linesDownload Raw Back to Text_TextNeedles
1business, but we may not be able to achieve those returns due to the factors discussed above. Additionally, the property and casualty 2insurance markets are historically cyclical and experience periods of relatively strong premium rates followed by periods of increased 3competition that push premium rates down. Inadequate pricing and the difference between estimated results of the above factors 4compared to actual results could have a material adverse effect on the profitability of our operations and our financial condition.5Guarantees within certain of our Life and Retirement products may increase the volatility of our results. 6Certain of our annuity and life insurance products include features that guarantee a certain level of benefits, including guaranteed 7minimum death benefits, guaranteed living benefits, including guaranteed minimum income benefits, and products with guaranteed 8interest crediting rates, including crediting rate guarantees tied to the performance of various market indices. Many of these features 9are accounted for at fair value as either MRBs or embedded derivatives under GAAP, and they have significant exposure to capital 10markets and insurance risks. An increase in valuation of liabilities associated with the guaranteed features results in a decrease in our 11profitability and depending on the magnitude of any such increase, could materially and adversely affect our financial condition, 12including our capitalization, as well as our financial strength ratings.13We employ a capital markets hedging strategy to partially offset the economic impacts of movements in equity, interest rate and credit 14markets, however, our hedging strategy may not effectively offset movements in our GAAP equity or our statutory surplus and capital 15requirements and may otherwise be insufficient in relation to our obligations. Furthermore, we are subject to the risk that changes in 16policyholder behavior or actual levels of mortality/longevity as compared to assumptions in pricing and reserving, combined with 17adverse market events, could produce losses not addressed by the risk management techniques employed. These factors, 18individually or collectively, may have a material adverse effect on our business, financial condition, results of operations or liquidity 19including our ability to receive dividends from our operating companies.20Changes in interest rates result in changes to the fair value liability. All else being equal, higher interest rates generally decrease the 21fair value of our liabilities, which increases our earnings, while low interest rates generally increase the fair value of our liabilities, 22which decreases our earnings. A prolonged low interest rate environment or a prolonged period of widening credit spreads may also 23subject us to increased hedging costs or an increase in the amount of statutory reserves that our insurance subsidiaries are required 24to hold for our liabilities, lowering their statutory surplus, which would adversely affect their ability to pay dividends. In addition, it may 25also increase the perceived value of our benefits to our policyholders, which in turn may lead to a higher than expected benefit 26utilization and lower than expected surrender rates of those products over time as compared to pricing assumptions.27Differences between the change in fair value of the GAAP MRBs and embedded derivatives, as well as associated statutory and tax 28liabilities, and the value of the related hedging portfolio may occur and can be caused by movements in the level of equity, interest 29rate and credit markets, market volatility, policyholder behavior and mortality/longevity rates that differ from our assumptions and our 30inability to purchase hedging instruments at prices consistent with the desired risk and return trade-off. In addition, we may sometimes 31choose not to hedge or fully mitigate these risks, based on economic considerations and other factors. The occurrence of one or more 32of these events has in the past resulted in, and could in the future result in, an increase in the fair value of liabilities associated with 33the guaranteed benefits without an offsetting increase in the value of our hedges, or a decline in the value of our hedges without an 34offsetting decline in our liabilities, thus reducing our results of operations and shareholders’ equity.35For additional information on these products, see Item 1. Business – Regulation, Part II, Item 7. MD&A – Critical Accounting 36Estimates – Market Risk Benefits and Notes 13 and 14 to the Consolidated Financial Statements.37Our risk management policies, standards and procedures may prove to be ineffective and leave us exposed to unidentified 38or unanticipated risk, which could adversely affect our businesses, results of operations, financial condition and liquidity. 39We have developed and continue to enhance enterprise-wide risk management policies, standards and procedures to identify, 40monitor and mitigate risk to which we are exposed. Our risk management policies, standards and procedures may not be sufficiently 41comprehensive and may not identify or adequately protect us from every risk to which we are exposed. Many of our methods of 42identifying, measuring, underwriting and managing risks are based upon our study and use of historical market, applicant, customer, 43employee and bad actor behavior or statistics based on historical models. As a result, these methods may not accurately predict 44future exposures from events such as a major financial market disruption as the result of a natural or man-made disaster (for 45example, a severe climate-related event or terrorist attack), that could be significantly different than the historical measures indicate, 46and which could also result in a substantial change in policyholder behavior and claims levels not previously observed. We have and 47will continue to enhance our underwriting processes, including, from time to time, considering and integrating newly available sources 48of data to confirm and refine our traditional underwriting methods. Our efforts at implementing these improvements may not, however, 49be fully successful, which may adversely affect our competitive position. We have also introduced new product features designed to 50limit our risk and taken actions on in-force business, which may not be fully successful in limiting or eliminating risk. We may take 51additional actions on our in-force business, including adjusting crediting rates and cost of insurance, which may not be fully successful 52in maintaining profitability and which may result in litigation. Moreover, our hedging programs and reinsurance strategies that are 53designed to manage market risk and mortality risk rely on assumptions regarding our assets, liabilities, general market factors and the 54creditworthiness of our counterparties that could prove to be incorrect or inadequate. Our hedging programs utilize various derivative 55instruments, including but not limited to equity options, futures contracts, interest rate swaps and swaptions, as well as other hedging 56ITEM 1A | Risk Factors57AIG | 2023 Form 10-K 27
AmazonScience/document-haystack · CoolFace