ALTK-Evolve: How learned memory tops the τ²-bench leaderboard on customer service tasks without a frontier model
A system using adaptive memory mechanisms outperformed larger proprietary models on customer service benchmarks, suggesting scale isn't the only path to performance.
Get smart on it
ALTK-Evolve is a technique that improves customer service agents by extracting behavioral guidelines from the agent's past attempts and adding those rules to its system prompt, without changing the model's weights. The approach uses an older model to both analyze what worked and didn't work in previous conversations and to power the agent itself, then applies the learned guidelines to new customer service tasks. This matters because it achieves top performance on a customer service benchmark while using a non-frontier model and a basic agent architecture, demonstrating that guidance added through prompting can significantly improve both accuracy and reliability. The technique was tested on two domains, retail and airline, where it improved Pass^4 reliability scores by up to 10 percentage points by teaching the agent to follow authentication procedures, confirm changes with customers, and execute actions systematically.
Explainer: How to Actually Read an AI Benchmark (Without Being Fooled)
Relevant courses
Want to go deeper than the news? Explore live, cohort-based AI courses taught by practitioners.
Browse AI courses on MavenRelated news
Inception says Mercury 3 beats GPT-6 Luna and Claude Haiku 5.5 at up to 8x the speed
Diffusion language model startup Inception AI Inc. today released Mercury 3, its new flagship model. In the company’s own benchmarks, Mercury 3 beat OpenAI Group PBC’s GPT-6 Luna and Anthropic PBC’s Claude Haiku 5.5 on eight of 12 tests at four to eight times their speed. The release takes on the main doubt about Inception’s […] The post Inception says Mercury 3 beats GPT-6 Luna and Claude Haiku 5.5 at up to 8x the speed appeared first on SiliconANGLE.
Deepseek V4.1 Flash 1017 beats larger rivals in agentic tasks while once again massively cutting per-task costs
Deepseek has released V4.1-Flash-1017 as an open-weight model that posts better benchmarks, finishes tasks faster, and produces shorter outputs than its predecessor. The company is also opening up its internal search engine through new Search and Fetch APIs at $0.75 per 1,000 calls. API users in tools like Claude Code and Codex get free access for now. The article Deepseek V4.1 Flash 1017 beats larger rivals in agentic tasks while once again massively cutting per-task costs appeared fir
Also reported by SiliconANGLE
Four Gemini 4 Argon agents beat Minecraft without any human help
Four AI agents running on Google's Gemini 4 Argon have beaten Minecraft for the first time without human help, slaying the Ender Dragon in a fully autonomous run. With no predefined roles or strategies, they coordinated on their own, wrote 2,783 programs, exchanged 229 chat messages, and died 605 times. The session cost $1,620. The article Four Gemini 4 Argon agents beat Minecraft without any human help appeared first on The Decoder.
Anthropics' Claude cracks a 25-year-old NSA cryptography mystery in under 30 minutes
Anthropic's AI model Claude has solved a 25-year-old cryptography puzzle. The mysterious seeds behind NIST's standard curves for digital signatures were provided by the NSA without explanation, raising long-standing suspicion about built-in backdoors. Claude figured out in under 30 minutes that the seeds came from "lrand48," a simple C random number generator, making deliberate backdoors far less likely. The article Anthropics' Claude cracks a 25-year-old NSA cryptography mystery i
Exa 3.0 Search API Tops Perplexity, Brave and Parallel on 8 Benchmarks
Exa has released Exa 3.0, an upgrade to its web search API for AI agents. The Exa 3.0 search API ships a larger embedding model and a cheaper search type, Exa Lite. Dynamic Highlights is now generally available. Exa claims that it leads Perplexity, Brave and Parallel on 8 benchmarks. Those benchmarks were run by Exa, so you can treat them as self-reported. For developers, the practical change is cost: Lite costs $4 per 1,000 requests. TL;DR Size: Exa tracks over 2 trillion URLs and serve
Elorian’s first AI model wants to reason with images, not just about them
According to Elorian's evaluations, Glimpse outperforms even GPT-6 Astra on scientific chart reasoning and short-form video captioning, with competitive results on other visual reasoning tasks.