Search “Search”

Imperial College London paper: search-augmented LLMs often cite web pages they never opened

Three authors from Imperial College London have published the paper "Retrieved but Not Read: Understanding Why Search-Augmented LLMs Fail to Use Available Evidence," examining why search-augmented LLMs frequently cite web pages they never actually open. The study tested GPT-5.6 Luna and six open-weight configurations including GLM-5.3-Flash, gpt-oss-120b, gpt-oss-20b, gemma-4-31b, and qwen3-8b using 225 facts that changed after February 16, 2026, finding that models often answer based solely on search snippets without opening the source pages; under the same set of search results, the proportion of cases where no page was opened ranged from 2.6% to 76%, and when Luna used its built-in search, 53.6% of turns relied only on snippets. When answering from snippets alone, outdated answers accounted for 6.4% and misleading citations for 6.4%, whereas when pages were opened and cited, these fell to 1.3% and 2.3% respectively; placing the correct page directly into the prompt significantly reduced errors, but some models still favored old information even when shown the new facts.

English translation is being prepared.

English translation is being prepared.

UC Santa Barbara paper proposes TTME: replacing test-time training with evolutionary memory for mathematical discovery

A UC Santa Barbara paper proposes TTME, which replaces test-time training with evolutionary memory in mathematical discovery tasks, keeping weights unchanged and storing rules and score-improvement ideas from previous attempts in memory to guide subsequent attempts. Across 5 mathematical tasks, 2 frameworks, and 3 models, TTME exceeded the reported test-time training scores every time; on 3 tasks with Qwen3-8B, its FLOPs were reduced by 82.8% to 84.6%, and runtime was reduced by 47.7% to 58.4%. The authors suggest that if an LLM search loop comes with an automatic checker, one can first try using experiential memory rather than paying the cost of weight updates in the middle of the search.

University of Texas paper proposes Jev-Mem: long-conversation memory construction 6.6x faster

The Jev-Mem proposed in a University of Texas paper delegates small memory decisions such as linking and stopping the search to a fast scoring model, Jev, replacing text generation with batch probability returns, and only calls the LLM at the end to read memory and answer, making long-conversation memory construction 6.6x faster. On the LoCoMo long-conversation benchmark, its answer score is 0.777, higher than the best baseline's 0.700; memory construction takes 158 seconds versus 1,044 seconds for the fastest comparison method, and answering takes 0.93 seconds versus 1.47 seconds for the comparison method. The authors suggest that if a memory pipeline calls an LLM for routing, scoring, or stop decisions, it can switch to a fast classifier and reserve the LLM for the final answer.

Convovo study: When forced to choose between two options, Grok 4.7 picks Republicans nine times out of ten, while GPT-6.1 Sol picks Democrats 97% of the time

A Convovo study of nearly 500 elections found that when forced to choose between two options via API, SpaceXAI's Grok 4.7 recommended the Republican candidate in 92% of responses, while OpenAI's GPT-6.1 Sol recommended the Democratic candidate in 97% of responses, with the two agreeing less than 10% of the time. The study also found that Grok frequently cited conservative sources, and that each API request carried about 1,240 additional tokens that researchers could neither view nor remove; after web search was disabled, GPT's Democratic lean persisted, while Grok's Republican lean disappeared, with 27 of 30 cases declining to choose. OpenAI responded that the test targeted the developer API and that the prompt explicitly required recommending a candidate, so it does not reflect the ChatGPT consumer experience, and that the consumer app would refuse to answer such questions.

W3C discussion draft proposes using Web standards to constrain AI agents

The W3C Strategy team published a discussion draft, "W3C and the State of the Web in the AI Era," proposing the use of Web standards to make AI agents safer, focusing on Prompt Injection protection, controllable permissions, and accessibility. The article was published on October 21 to build momentum for the TPAC meeting; it is a proposal rather than an approved standard, with the official presentation scheduled for October 29. Using travel booking as an example, the article points out that letting an agent search for flights does not equal authorizing it to pay for tickets, and recommends setting spending limits, revocable authorization, and traceable credentials for payments.

Librarian Agents paper: letting agents organize long contexts themselves to improve retrieval performance

Chelsea Finn retweeted a new paper, Librarian Agents, by Anikait Singh and others, which studies whether agents can adapt to large amounts of unstructured context in specialized domains by organizing their own long-context workspaces and anticipating queries. She notes that agentic search can scale to information of any size but tends to miss relevant content, while optimizing information organization for downstream use can also bring 30+% improvements in OOD domains. The tweet includes links to the paper and code.

Microsoft Research proves the 31-year-old unsolved KLS conjecture

Microsoft Research researcher Xinzhi Zhang and co-authors proved the KLS conjecture proposed in 1995, giving a dimension-independent fixed bound; the previous best result only allowed the bound to grow slowly with dimension. The proof came from a human-guided agent search process, in which the team explored more than 100 candidate paths, with researchers evaluating, pruning, and redirecting them; an independent team also gave a complete proof at the same time, followed by Lean 4 formal verification. The authors say earlier unguided end-to-end attempts with multiple frontier models failed to achieve the goal.

English translation is being prepared.

Anthropic has Claude autonomously design quantum circuits, with some results outperforming published approaches

Anthropic gave Claude plain-language specifications for eight quantum circuits along with an automated evaluator; in a closed loop with essentially no human intervention, Claude repeatedly submitted designs and improved them based on scores, finding solutions that use fewer qubits and gates than the best published circuits, with some improvements reaching an order of magnitude. The elliptic-curve modular inversion design came from a single 53-hour continuous run, in which Claude derived a reversible divsteps implementation that requires no storage of branch information and formally verified it with Lean; another effort reduced the gate count of the most expensive Reed-Solomon decoding step in DQI to about 1/12 of Google's previous construction. Anthropic also noted that Claude was initially conservative and tended to stick with known methods, which had to be mitigated by requiring parallel exploration of multiple approaches and occasionally prompting it to broaden its search.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.