Saturated benchmarks are losing discriminative power, while multilingual tests reveal that English-centric LLM performance does not transfer to languages like Bengali. New evaluation methods, including C-tests and transcript mining, are emerging to address these blind spots.
When several models reach near-perfect scores, benchmarks lose their ability to differentiate them. A study shows that full task transcripts from saturated benchmarks retain predictive information. Pairwise judging reduced prediction error by 5.3 tim...
A zero-shot benchmark of eight LLMs shows that summarization quality does not transfer from English to Bengali. GPT-5.6 Luna leads several English lexical metrics, but LLaMA 3.3 70B achieves the best Bengali ROUGE and BLEU scores; Gemini 2.5 Flash po...
A new test uses Levin's Kt and language models as representational machines to generate items whose difficulty is measurable in bits, before any human or frontier model can solve them. Twelve LLMs were evaluated on a first certified battery, showing ...
