It’s kind of weird how during the island hopping campaign, the Japanese would fight to the last man and marines were forced to fight for every square foot of ground on Saipan, Iwo Jima, and Okinawa but once a bunch of T-34s roll into Manchuria they just throw in the towel.
The large-scale surrenders in Manchuria were after the Japanese surrender of 15th August. The simple version is that they threw in the towel because they'd been ordered to throw in the towel, by the Emperor. They didn't throw in the towel just because a bunch of T-34s rolled into Manchuria (even when said bunch was well over 3000 tanks and almost 2000 self-propelled guns). Resistance continued after that date in many places, for example at Sunwu where the defenders surrendered 18th August, and Ainhu 20th August.
While Japanese resistance collapsed in many places, this didn't result in mass surrenders of Japanese soldiers. The small number of prisoners before the final surrenders is remarkable considering that the Japanese were outnumbered about 2 to 1, faced an overwhelming Soviet advantage in armour (Soviet armour was much more numerous, much better quality, and the Japanese severely lacking in anti-tank weapons), artillery, and air, and were, in many places, attacked by surprise in unfortified positions.
Compared to the island battles in the Pacific, the Soviets could often bypass strongly defended areas (and needed to do so, to maintain their planned rapid advance), and the Japanese could retreat. This meant that the fighting didn't always need to be a frontal grind through defences such as forced by the limited space in many of the island battles. Add to that the Japanese surrender on 15th August, less than a week after the Soviet invasion - there were plenty of Japanese forces still in the field, able to put down their arms at the end. These were the vast majority of the Japanese soldiers in Soviet hands at the end of the war.
The Soviet forces suffered about 12,000 KIA, while the Japanese may have suffered up to over 80,000. Just over 80,000 is the Soviet estimate, while the official Japanese estimate of just over 20,000 is certainly an underestimate since that is the number of KIA recorded by the Japanese army, in records that survived the end of the war. The Soviet estimate is probably closer to the true figure than Japanese estimate is. The determination of the Japanese forces is reflected by these casualty figures, and by the successful resistance in some regions, for almost a week after the Japanese surrender. Further, a large part of the tank losses (over 300) suffered by the Soviet forces were inflicted by suicide bombers (infantry with anti-tank bombs strapped to themselves) - not the act of an army throwing in the towel.
David Glantz wrote an excellent analysis of the Soviet invasion of Manchuria; this has been published as a book, but is also freely available online as a US Army study: