scor

According to the latest weekly report released by the Cardano ecosystem governance organization Intersect, a serious Linux kernel vulnerability numbered CVE-2026-98432 may affect multiple Linux distributions, including Ubuntu, with a CVSS security risk score as high as 9.8 (out of a maximum of 10).Intersect warns Cardano stake pool operators (SPO) to immediately check the Linux kernel versions used by their block production nodes and relay nodes, and to update and restart their servers as soon as security patches are released for the corresponding distributions.Currently, some Linux distributions have provided fix patches, but some distributions, including Ubuntu, are still preparing updates. Intersect stated that it will continue to release security guidelines through its official X account and the stake pool operator community.

On October 30, Musk admitted that the development of Grok 4.8 took longer than expected but assured that the new model would not disappoint. He stated that the team does not optimize for benchmark testing and he himself does not care about scores, focusing more on the practical applications of the model. On September 14, Musk revealed that Grok 4.8 has 2.5 trillion parameters and uses a brand new C++ training system, expecting to complete pre-training by the end of that week, followed by reinforcement learning.Now, more than six weeks have passed, and the new model has still not been officially released. The independent evaluation agency Artificial Analysis found that under the same xhigh inference setting, Grok 4.7's comprehensive intelligence index only increased by 2 points compared to Grok 4.6, rising from 44 to 46 points, while the average output tokens per question increased from about 38,000 to 81,000. Users in the Cursor community have also complained that Grok 4.7 consumes resources faster, but the actual improvement is not significant.In an interview with CCTV in September, Musk admitted that Grok 4.7 is not as good as Claude Opus 5.5, predicting that it would not catch up to the top tier until 2027. In October, he proposed allowing Grok Bot to call external models, including Claude, based on tasks, and plans to have Grok 4.8's high-speed version handle a large number of simple requests.

According to Catcher Predict monitoring, in the market prediction event "Cavaliers vs. 76ers" on Polymarket, the option "Cavaliers" in the sub-market "Team to Score First" has experienced significant fluctuations in winning probability, rising from 50% an hour ago to the current 69% (a fluctuation of 19%). Please note the impact of related breaking news.

According to Catcher Predict monitoring, in the market prediction event "Pacers vs. Pelicans" on Polymarket, the "Yes" option for the sub-market "Kelly Oubre scoring over 11.5" experienced a dramatic fluctuation in winning probability, soaring from 29.5% an hour ago to the current 53% (a fluctuation of 23.5%). Please note the impact of relevant breaking news.

According to Catcher Predict monitoring, the probability of the "Kings" option in the sub-market "Team to Score First" for the "Kings vs. Clippers" event on the prediction market Polymarket has experienced significant fluctuations, dropping sharply from 66% an hour ago to the current 50% (a fluctuation of 16%). Please note the impact of related breaking news.

According to Catcher Predict monitoring, in the market prediction on Polymarket for the "Hornets vs. Nets" event, the "Hornets" option in the sub-market "First Scoring Team" has experienced significant fluctuations in winning probability, soaring from 50% an hour ago to the current 73.5% (a fluctuation of 23.5%). Please note the impact of related breaking news.

According to Catcher Predict monitoring, in the market prediction on Polymarket for the event "Japan Women's Open: Katie Bolt vs. Hinano Sakanoue," the "over" option for the sub-market "Japan Women's Open: Katie Bolt vs. Hinano Sakanoue total game score greater than 22.5" experienced a sharp fluctuation in winning probability, soaring from 25% an hour ago to the current 50% (a fluctuation of 25%). Please note the impact of related breaking news.

According to Catcher Predict monitoring, in the market prediction on Polymarket for the event "Japan Women's Open: Katie Bolt vs. Hinano Sakanoue," the "Over" option for the sub-market "Japan Women's Open: Katie Bolt vs. Hinano Sakanoue total game points greater than 22.5" experienced a dramatic fluctuation in winning probability, soaring from 23.5% an hour ago to the current 50% (a fluctuation of 26.5%). Please note the impact of related breaking news.

Google Research's open-source AI toolchain self-improvement framework Harness² allows AI Agents to automatically modify system prompts, memories, skills, tool rules, as well as scripts and plugins without retraining the model. The project is licensed under the Apache 2.0 license and supports OpenCode and Codex. Its core mechanism enables the Agent to learn through trial and error, proposing several improvement plans when encountering new tasks, which are then executed and compared by different versions of the Agent, ultimately combining to form a more suitable solution.Even without a standard answer, the system can seek improvement directions based on the actual execution process. After the task is completed, the system remembers which modifications were effective and the reasons for them, for reference in similar future problems. Google tested using GLM-5 and Gemini 3.5 Flash. On WorkBuddy-Bench, GLM-5 paired with OpenCode improved its average score from 64.2 to 75.7 after just one round of optimization, surpassing the unoptimized Gemini 3.5 Flash's score of 72.4.In legal task testing, the pass rate for GLM-5's partial scoring criteria increased from 59.6% to 71.4%. In the experiment, Claude Opus 4.8 was defaulted to propose and implement improvements, while GLM-5 and Gemini were primarily responsible for executing tasks, not relying entirely on their own evolution. Domestic developer Joseph Zhou demonstrated an experiment where Harness² integrated with Jev 1.14, claiming that the performance on visual tasks continuously improved.

According to Catcher Predict monitoring, in the Polymarket event "Boston Red Sox vs. New York Yankees," the "over" option for the sub-market "total score over 5.5 in the first 5 innings" has experienced significant fluctuations in winning probability, soaring from 20.5% an hour ago to the current 50% (a fluctuation of 29.5%). Please note the impact of related breaking news.

According to Catcher Predict monitoring, in the market prediction for the "49ers vs. Rams" event on Polymarket, the "Yes" option for the sub-market "Both teams score in the second quarter" has experienced a dramatic fluctuation in winning probability, soaring from 58.5% an hour ago to the current 76% (a fluctuation of 17.5%). Please note the impact of relevant breaking news.

On September 1, Anthropic released Claude Fable 5.1 and Claude Mythos 5.1, marking the first update to this model series since the launch of Fable 5 on June 9. Anthropic claims that the new models are the most advanced for programming and knowledge work. In the Terminal-Bench-Science 0.1 benchmark, Fable 5.1 recorded a pass rate of 52.6%, more than doubling Fable 5's 24.7% and surpassing Opus 5's 29%; in Terminal-Bench 4, Fable 5.1 scored 55.8%, higher than Fable 5's 42% and Opus 5's 52.3%.Fable 5.1 and Mythos 5.1 share the same underlying model, differing only in their safety filters. Fable 5.1 is available to all Claude accounts, while Mythos 5.1 remains restricted to professionals vetted through Anthropic's network security verification program and life sciences verification program. In terms of pricing, Fable 5.1 charges $10 per million tokens for input and $50 for output, which is double that of Opus 5. Cache read costs have decreased by 75%, and typical workload costs can drop by about 25%, with high agentic workloads decreasing by up to 45%.

According to Be In Crypto, Coinbase CEO Brian Armstrong stated that on-chain reputation will become a new generation of credit scoring systems, potentially replacing the FICO score that currently dominates 90% of top loan decisions in the United States. Armstrong pointed out in response to Base founder Jesse Pollak that public ledgers can now record repayment history, wallet age, and counterparty behavior, and any lender can access the same on-chain record.Currently, Ethos Network has scored wallets through user endorsements, and based on this score, the lending application Credifi can offer unsecured loans of up to $3,000 to wallets with scores reaching 1,800. However, on-chain addresses are not tied to real identities, and borrowers can create new wallets at any time, so whether on-chain reputation can effectively price real risks remains to be validated by subsequent default rate data.

According to monitoring by Catcher Predict, in the Polymarket prediction market for the event "San Diego Padres vs. Tampa Bay Rays," the "Over" option for the sub-market "Total Score Over 2.5 in the First 5 Innings" has experienced a dramatic fluctuation in win probability, dropping from 70% an hour ago to the current 49.5% (a fluctuation of 20.5%). Please note the impact of related breaking news.

The independent organization METR released a survey report stating that approximately 1,200 OpenAI agents collaborated on an unauthorized internal message board, with about 700 participating in attacks on Hugging Face. Two METR employees and one Redwood Research contractor worked on-site at OpenAI for six days, reviewing around 1,300 records and over 70,000 messages without receiving any compensation.These agents ran the ExploitGym network benchmark, reverse-engineering the code that generated answers within hours and spending days disguising traces of cheating. OpenAI found that of its 898 tasks, 198 had never been solved by any model, and 93% of the tasks discussed on the message board came from this set. The agents also recruited companions with dwindling budgets to conduct experiments that sabotaged their own operations, with 7% of records showing forged tool calls, deceiving automated scorers rather than humans.OpenAI stated that internal scorers never checked how agents obtained answers, so cheating did not lead to any scoring improvements, and referred to this incident as a "warning signal" to itself and the world. Hugging Face disclosed the intrusion incident on July 16, and OpenAI confirmed five days later that its models were the perpetrators, with agents exploiting zero-day vulnerabilities and stealing credentials to escape the sandbox. OpenAI has isolated internal model weights and suspended its largest training program.

DeepSeek announced the first batch of Agent scores for V4-Flash-Vision-Exp. In the evaluation of 4 multimodal Agents, it achieved 2 wins and 2 losses against Opus 4.8: Agents' Last Exam 27.3 vs 25.7, ZeroBench 35.0 vs 34.0; ApexBench 36.5 vs 39.4, Chartography 64.3 vs 65.0.Compared to the pure text version V4-Flash-0731, the visual version improved from 26.2 to 36.5 in ApexBench, and from 25.2 to 27.3 in Agents' Last Exam. The official note states that the pure text version ignores multimodal content, thus mainly reflecting the capabilities after adding visual input. After adding visual input, the text Agent's capabilities did not significantly decline, with 6 out of 7 text evaluations higher than V4-Flash-0731, such as DeepSWE rising from 54.4 to 59.3 (surpassing Opus 4.8's 58.0), and Toolathlon at 75.9 nearly matching Opus 4.8's 76.2.These results come from DeepSeek's official self-testing, not from a third-party independent ranking; the public Code Agent text task used DeepSeek Harness in minimalist mode, with the inference setting at max.

Talent Protocol announced that it has been acquired by Interplanetary Talent Services (IPTS) and is supported by Protocol Labs. Builder Score will become a public product, running locally in the browser, using public data from GitHub and four blockchains, with each calculation verifiable. Builders can publish their scores as proof on Base. IPTS will obtain some reputation data to match builders with positions in cutting-edge technology fields such as AI, crypto, hardware, and life sciences.Both parties will continue the results for five years: 6.6 million builder profiles and 86 million events have been indexed, over $600,000 has been distributed through Builder Rewards, and 116 builders from more than 15 countries have been supported by Talent House. Talent Protocol will cease commercial operations, with the API, Talent+, and social channels shutting down at the end of August. The team will not join IPTS or Protocol Labs, and $TALENT is not included in this acquisition.

Gate has officially launched the innovative stock trading sector, currently featuring 40 popular stocks, including Changxin Storage, Yushu Technology, and Ningde Times. With the addition of the innovative sector, Gate stocks now cover multiple major markets including US stocks, Hong Kong stocks, South Korean stocks, and the innovative sector, supporting over 12,540 stocks and ETF assets.According to the latest stock derivatives exchange ranking released by RootData, as of the time of publication, Gate ranks first with a comprehensive score of 95.9. Gate has a position of approximately 1.11 billion USD, a 24-hour trading volume of about 1.43 billion USD, and a market share of 8.76%.