🌱 Seed-2.1-Pro Review: Better Post-Training Can't Hide an Aging Base @ByteDanceSeed_ Seed-2.1-Pro 0915 earns higher reasoning scores with fewer tokens, and its agent ability has climbed from nearly unusable to passable. But in the three months since the last version, rivals iterated roughly a generation and a half — and the old base model is running out of tricks. That is the verdict from Zhihu contributor toyama nao, who runs a long-running monthly logic benchmark and put the 0915 build through his full evaluation suite. 1️⃣ Coding and agent work: from unusable to passable The jump over the predecessor is large. Deliverable completeness now beats DeepSeek V4.1 Flash, though first-pass success still trails top-tier models — a model caught in the middle. 🔹 Frontend: some aesthetic sense, but unstable. Constrained stacks like iOS system components look decent; open stacks like web or raw canvas expose visible flaws in proportion and color. 🔹 Task adaptability: every task now at least completes. In a HarmonyOS-flavored project it barely knew, the model read docs and iterated its way to usable — a good sign. 🔹 Delivery efficiency: roughly tied with DeepSeek V4.1 Flash and GLM-5.3-Flash, splitting steps evenly between writing and verifying. All three sit below the current Chinese SOTA. 2️⃣ Multi-step reasoning: the biggest gain This is where 0915 improved most — and where it leads its tier, with a small token-efficiency edge. On problems where GLM-5.3 falls into exhaustive enumeration, 0915 repeatedly finds higher-scoring answers with fewer tokens, showing what the author calls real "big-model intuition." The caveat: multi-turn reasoning, which demands in-context learning and reflection, remains mediocre — on par with Chinese peers. 3️⃣ Where 0915 still stumbles Hallucination runs high, and context confusion appears regardless of prompt length — especially when source details are tangled. The agent symptom is subtler than dropping requirements: 0915 keeps every requirement but misreads semi-ambiguous ones — arguably the more dangerous failure mode. 4️⃣ The contrarian bet: a true non-thinking mode Seed is one of the few teams still maintaining a genuinely non-thinking mode, and this version quietly got good: average output dropped from 8K tokens back to ~1K, with no measurable capability regression, a slight gain in complex reasoning, and readable prose intact. For latency-sensitive scenarios that still need some reasoning, the author considers it a legitimate option. 5️⃣ The long march His closing frames 0915 as a rest stop, not a destination: the predecessor's lukewarm market reception forced ByteDance's team into a forced march of biweekly iterations, and better post-training is now visibly paying off in cost per task. But the base is old, and the competition has moved. His last line is worth keeping: sometimes the long way around is the real shortcut. 🔗 Full Reading: 🔗 Key links: Author's monthly logic benchmark (Aug 2026): #ByteDance# #Seed# #LLM# #AIAgents# #LLMBenchmark# #ReasoningModels# #AI#

TRUE ROMANCE. 1993. Tarantino wrote it. Tony Scott shot it. The poster hangs on my studio wall today. One of my favorite films. Clarence at the sink. The gold jacket over his shoulder, half a face in the mirror, talking low. "I like you… always have, always will." Kilmer isn't even credited as Elvis — the estate wouldn't have it, so the credits just say "Mentor." Eight hours in the makeup chair for two days of work. A mirror, a silhouette, and a voice. Best cameo of the decade. And Scott never lets the violence off the hook. Every bullet in that picture has a receipt attached. Blood on the wallpaper, a father in a chair, a kid who wanted to be Elvis. That's the job — make it cost something. Thirty-three years on. Still holds. What's the scene that stuck to you? #TrueRomance# #TonyScott# #ValKilmer# #Tarantino# #FilmCraft# #IrishChannelStudios#

Obamacare was "Affordable Health Care" The Democrats insisted this was true when ONLY Democrats voted in favor and forced it down the throats of the American people. They don't get to do it again. Obamacare has almost broken the system completely.

One of our favorite ways to explain zero-knowledge proofs: 🪪Showing your ID at a bar proves the one thing the bouncer needs, that you're over 21. But it also reveals everything else: your full name, exact birthday, address, ID number. A zero-knowledge proof is the version where you prove only the part that matters, "over 21," and nothing else. That's one of the core ideas: prove something is true without exposing the data behind it. And it goes way past age checks: 🔒 prove you can afford a loan without revealing your balance ⚙️ prove a computation ran correctly without revealing its inputs What's your favorite analogy for explaining ZK?

Going direct is the best comms approach today - you get the true authentic message out there and connect properly with your users. There’s no people in the middle twisting things, and you hear direct feedback from the people you care about. Do yourself and your company a favor, go direct more often!