Playing with DeepSeek-R1 in LM Studio helps to show why this is the case. Reading it's "thought" process is intriguing and helps to show where possible errors occurred in the provided answers. Now, if DeepSeek could take the Perplexity feature of providing links to the sourced information, it'd be damn near perfect!
this post was submitted on 27 Jan 2025
11 points (100.0% liked)
United States | News & Politics
7481 readers
377 users here now
founded 4 years ago
MODERATORS