iktHow to centre a div
marcos
TeddEJust as a tangent:
This is one reason why I’ll never trust AI.
I imagine we might wrangle the hallucination thing (or at least be more verbose about it’s uncertainty), but I doubt it will ever identify a poorly chosen question.
Making the LLMs warn you when you ask a known bad question is just a matter of training it differently. It’s a perfectly doable thing, with a known solution.
Solving the hallucinations in LLMs is impossible.
LeonThat’s because it’s a false premise. LLMs don’t hallucinate, they do exactly what they’re meant to do; predict text, and output something that’s legible and human written. There’s no training for correctness, how do you even define that?
There’s no training for correctness, how do you even define that?
I guess can chat to these guys who are trying:
By scaling reasoning with reinforcement learning that rewards correct final answers, LLMs have improved from poor performance to saturating quantitative reasoning competitions like AIME and HMMT in one year
Sure, when it comes to mathematics you can do that with extreme limitations on success, but what about cases where correctness is less set? Two opposing statements can be correct if a situation changes, for example.
The problems language models are expected to solve go beyond the scope of what language models are good for. They’ll never be good at solving such problems.
The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem Complexity
Recent generations of language models have introduced Large Reasoning Models (LRMs) that generate detailed thinking processes before providing answers. While these models demonstrate improved performance on reasoning benchmarks, their fundamental capabilities, scaling properties, and limitations remain insufficiently understood. Current evaluations primarily focus on established math and coding benchmarks, emphasizing final answer accuracy. However, this evaluation paradigm often suffers from contamination and does not provide insights into the reasoning traces. In this work, we systematically investigate these gaps with the help of controllable puzzle environments that allow precise manipulation of complexity while maintaining consistent logical structures. This setup enables the analysis of not only final answers but also the internal reasoning traces, offering insights into how LRMs think. Through extensive experiments, we show that LRMs face a complete accuracy collapse beyond certain complexities. Moreover, they exhibit a counterintuitive scaling limit: their reasoning effort increases with problem complexity up to a point, then declines despite having remaining token budget. By comparing LRMs with their standard LLM counterparts under same inference compute, we identify three performance regimes: (1) low-complexity tasks where standard models outperform LRMs, (2) medium-complexity tasks where LRMs demonstrates advantage, and (3) high-complexity tasks where both models face complete collapse. We found that LRMs have limitations in exact computation: they fail to use explicit algorithms and reason inconsistently across scales. We also investigate the reasoning traces in more depth, studying the patterns of explored solutions and analyzing the models' computational behavior, shedding light on their strengths, limitations, and raising questions about their reasoning capabilities.
i duno you’re in the wrong forum, you want hackernews or reddit, no one here knows much about ai
although you do seem to be making the same mistake others made before where you want point to research happening currently and then extrapolating that out to the future
ai has progressed so fast i wouldn’t be making any “they’ll never be good at” type statements