The Last Responsible Moment: All Three Models Passed—So What Did the Benchmark Actually Measure?
This is a submission for the Kaggle Benchmarking Challenge What I Benchmarked A coding...
#Benchmark
1661 articles tagged with Benchmark
I Built a Benchmark That Catches AI Models Cheating (And They All Failed)
I Built a Benchmark That Catches AI Models Cheating (And They All Failed) What...
AI agent benchmark: I gave 9 models a destroy button and a job that needed it
This is a submission for the Kaggle Benchmarking Challenge An AI agent benchmark for tool use should...
Can AI tell whether an optimum is unique? A small Kaggle benchmark with exact answers
This is a submission for the Kaggle Benchmarking Challenge What I Benchmarked When a...
Beyond Leaderboard Illusions: Benchmarking Multi-Turn Agentic Feedback Loops in Autonomous Software Engineering with TFD-Bench
Beyond Leaderboard Illusions: Benchmarking Multi-Turn Agentic Feedback Loops in Autonomous...
📊 Olmo 3 7B Think — the actual numbers GPQA: 51.6% MMLU-Pro: 65.5% Humanity's Last Exam: 6% Long Context Reasoning: 0%Me...
📊 Olmo 3 7B Think — the actual numbers GPQA: 51.6% MMLU-Pro: 65.5% Humanity's Last Exam: 6% Long Context Reasoning: 0%Measured independently, not self-reported →https://olud.ai/lea...
I ported my coding-agent benchmark to Kaggle, and the first bugs I found were mine
This is a submission for the Kaggle Benchmarking Challenge What I Benchmarked I maintain...
The Agent Finished the Job. The Benchmark Gave It Zero.
An agent submits a support ticket. The business system creates it, but the connection breaks before...
Last Secure Code Benchmark: Is the Code You Vibe Secure?
AI coding agents now write much of the code that ships, yet whether that code is secure remains unclear. We introduce the Last Secure Code Benchmark, a comprehensive evaluation of ...
LIQUIDITY EVENT IDENTIFICATION An LLM Reasoning Benchmark Kaggle Benchmarking...#ai #llm #machinelearning #software #cod...
LIQUIDITY EVENT IDENTIFICATION An LLM Reasoning Benchmark Kaggle Benchmarking...#ai #llm #machinelearning #software #coding #development #engineering #inclusive #community[Boost]
Let's check characters. Write: "Kimi K3 scores 43.6 vs Claude Fable 5.1's 53.4 — a 9.8-pointhttps://olud.ai/leaderboard....
Let's check characters. Write: "Kimi K3 scores 43.6 vs Claude Fable 5.1's 53.4 — a 9.8-pointhttps://olud.ai/leaderboard.html#OpenSource #AI #LLM
osfv/talus: Talus: a 23M-parameter diffusion model that generates game terrain heightmaps. Weights, training and benchmark code (Apache-2.0).
Talus: a 23M-parameter diffusion model that generates game terrain heightmaps. Weights, training and benchmark code (Apache-2.0).
📦 Fresh on the leaderboard: GPT-6.1 Sol Pro (OpenAI)Context: 1.1M tokens · $2 in / $10 out per 1M · commercialAll the la...
📦 Fresh on the leaderboard: GPT-6.1 Sol Pro (OpenAI)Context: 1.1M tokens · $2 in / $10 out per 1M · commercialAll the latest models, tracked hourly:https://olud.ai/latest.html#AI #...
📊 Arena, leaderboard per valutare gli AI, vale 3,1 miliardiArena (ex LMSYS), nata all'UC Berkeley nel 2023, ha chiuso un...
📊 Arena, leaderboard per valutare gli AI, vale 3,1 miliardiArena (ex LMSYS), nata all'UC Berkeley nel 2023, ha chiuso una Serie B da 00M a ,1B di valuation. Il sito, usato da milio...
Researchers introduce a verification-centric perspective and a scalable benchmark for LLM-assisted peer review, emphasiz...
Researchers introduce a verification-centric perspective and a scalable benchmark for LLM-assisted peer review, emphasizing error detection.Source: arXivhttps://arxiv.org/abs/2610....
Researchers introduced PARCEL, a new benchmark designed to detect legal hallucinations in large language models by check...
Researchers introduced PARCEL, a new benchmark designed to detect legal hallucinations in large language models by checking if claims are supported by cited sources.Source: arXiv c...
A measurement view of the System One Mosaic Benchmark, where a zero-token judge leads 102 models across 137 tasks by ran...
A measurement view of the System One Mosaic Benchmark, where a zero-token judge leads 102 models across 137 tasks by rank-based Borda scoring.#ai #machinelearning #benchmarks #data...
The public leaderboard says Model X is #1. Your production traffic disagrees. Here’s how to build the...#ai #machinelear...
The public leaderboard says Model X is #1. Your production traffic disagrees. Here’s how to build the...#ai #machinelearning #llm #architecture #software #coding #development #engi...
My benchmark scored GPT-5.4 mini 1.00 by silently skipping the 3 questions it failed
This is a submission for the Kaggle Benchmarking Challenge GPT-5.4 mini got a perfect 1.00 on my...
Show HN: AI SRE Arena, an Open Benchmark for AI SRE Agents on KubernetesArticle URL: https://github.com/edgedelta/projec...
Show HN: AI SRE Arena, an Open Benchmark for AI SRE Agents on KubernetesArticle URL: https://github.com/edgedelta/project-arena Comments URL: https://news.ycombinator.com/item?id=5...
AgentHorizon is a benchmark of 1,373 computer-use tasks designed to evaluate agentic judges for long-horizon tasks.Sourc...
AgentHorizon is a benchmark of 1,373 computer-use tasks designed to evaluate agentic judges for long-horizon tasks.Source: arXivhttps://arxiv.org/abs/2610.11050#MachineLearning
How an eval set is built, and why it rots
An eval set is an instrument with a purpose, a scoring rule, a calibration record and an expiry date. Where the samples come from, what counts as correct, and what makes the set ex...
LLMs Pass the Data-Science Quiz, Then Give Different Advice: A Kaggle Benchmark of 36 Measured Judgment Calls
This is a submission for the Kaggle Benchmarking Challenge. What I Benchmarked I spend a...
LLMs Pass the Data-Science Quiz, Then Give Different Advice: A Kaggle Benchmark of 36 Measured Judgment Calls#devchallen...
LLMs Pass the Data-Science Quiz, Then Give Different Advice: A Kaggle Benchmark of 36 Measured Judgment Calls#devchallenge #kagglechallenge #machinelearning #ai #software #coding #...