Be a part of our each day and weekly newsletters for the most recent updates and unique content material on industry-leading AI protection. Be taught Extra
DeepSeek-R1 has certainly created quite a lot of pleasure and concern, particularly for OpenAI’s rival mannequin o1. So, we put them to check in a side-by-side comparability on just a few easy information evaluation and market analysis duties.
To place the fashions on equal footing, we used Perplexity Professional Search, which now helps each o1 and R1. Our purpose was to look past benchmarks and see if the fashions can truly carry out advert hoc duties that require gathering data from the online, choosing out the proper items of knowledge and performing easy duties that might require substantial guide effort.
Each fashions are spectacular however make errors when the prompts lack specificity. o1 is barely higher at reasoning duties however R1’s transparency offers it an edge in instances (and there will likely be fairly just a few) the place it makes errors.
Here’s a breakdown of some of our experiments and the hyperlinks to the Perplexity pages the place you possibly can evaluate the outcomes your self.
Calculating returns on investments from the online
Our first check gauged whether or not fashions may calculate returns on funding (ROI). We thought-about a state of affairs the place the consumer has invested $140 within the Magnificent Seven (Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia, Tesla) on the primary day of each month from January to December 2024. We requested the mannequin to calculate the worth of the portfolio on the present date.
To perform this activity, the mannequin must pull Magazine 7 value data for the primary day of every month, break up the month-to-month funding evenly throughout the shares ($20 per inventory), sum them up and calculate the portfolio worth in line with the worth of the shares on the present date.
On this activity, each fashions failed. o1 returned an inventory of inventory costs for January 2024 and January 2025 together with a system to calculate the portfolio worth. Nevertheless, it didn’t calculate the proper values and mainly mentioned that there can be no ROI. Then again, R1 made the error of solely investing in January 2024 and calculating the returns for January 2025.
Nevertheless, what was fascinating was the fashions’ reasoning course of. Whereas o1 didn’t present a lot particulars on the way it had reached its outcomes, R1’s reasoning traced confirmed that it didn’t have the proper data as a result of Perplexity’s retrieval engine had didn’t get hold of the month-to-month information for inventory costs (many retrieval-augmented technology purposes fail not due to the mannequin lack of skills however due to unhealthy retrieval). This proved to be an necessary little bit of suggestions that led us to the following experiment.
Reasoning over file content material
We determined to run the identical experiment as earlier than, however as a substitute of prompting the mannequin to retrieve the data from the online, we determined to supply it in a textual content file. For this, we copy-pasted inventory month-to-month information for every inventory from Yahoo! Finance right into a textual content file and gave it to the mannequin. The file contained the title of every inventory plus the HTML desk that contained the worth for the primary day of every month from January to December 2024 and the final recorded value. The info was not cleaned to scale back the guide effort and check whether or not the mannequin may choose the proper elements from the info.
Once more, each fashions failed to supply the proper reply. o1 appeared to have extracted the info from the file, however instructed the calculation be executed manually in a software like Excel. The reasoning hint was very imprecise and didn’t comprise any helpful data to troubleshoot the mannequin. R1 additionally failed and didn’t present a solution, however the reasoning hint contained quite a lot of helpful data.
For instance, it was clear that the mannequin had accurately parsed the HTML information for every inventory and was capable of extract the proper data. It had additionally been capable of do the month-by-month calculation of investments, sum them and calculate the ultimate worth in line with the most recent inventory value within the desk. Nevertheless, that last worth remained in its reasoning chain and didn’t make it into the ultimate reply. The mannequin had additionally been confounded by a row within the Nvidia chart that had marked the corporate’s 10:1 inventory break up on June 10, 2024, and ended up miscalculating the ultimate worth of the portfolio.
Once more, the actual differentiator was not the outcome itself, however the potential to research how the mannequin arrived at its response. On this case, R1 offered us with a greater expertise, permitting us to grasp the mannequin’s limitations and the way we will reformulate our immediate and format our information to get higher outcomes sooner or later.
Evaluating information over the online
One other experiment we carried out required the mannequin to check the stats of 4 main NBA facilities and decide which one had the very best enchancment in area purpose proportion (FG%) from the 2022/2023 to the 2023/2024 seasons. This activity required the mannequin to do multi-step reasoning over completely different information factors. The catch within the immediate was that it included Victor Wembanyama, who simply entered the league as a rookie in 2023.
The retrieval for this immediate was a lot simpler, since participant stats are broadly reported on the internet and are normally included of their Wikipedia and NBA profiles. Each fashions answered accurately (it’s Giannis in case you have been curious), though relying on the sources they used, their figures have been a bit completely different. Nevertheless, they didn’t notice that Wemby didn’t qualify for the comparability and gathered different stats from his time within the European league.
In its reply, R1 offered a greater breakdown of the outcomes with a comparability desk together with hyperlinks to the sources it used for its reply. The added context enabled us to appropriate the immediate. After we modified the immediate specifying that we have been searching for FG% from NBA seasons, the mannequin accurately dominated out Wemby from the outcomes.
Ultimate verdict
Reasoning fashions are highly effective instruments, however nonetheless have a methods to go earlier than they are often absolutely trusted with duties, particularly as different parts of huge language mannequin (LLM) purposes proceed to evolve. From our experiments, each o1 and R1 can nonetheless make primary errors. Regardless of exhibiting spectacular outcomes, they nonetheless want a little bit of handholding to provide correct outcomes.
Ideally, a reasoning mannequin ought to be capable of clarify to the consumer when it lacks data for the duty. Alternatively, the reasoning hint of the mannequin ought to be capable of information customers to higher perceive errors and proper their prompts to extend the accuracy and stability of the mannequin’s responses. On this regard, R1 had the higher hand. Hopefully, future reasoning fashions, together with OpenAI’s upcoming o3 sequence, will present customers with extra visibility and management.
