@DeadHat
I 'sorta' make bots. They're all over the place. Some are niche. Most are odd. I mostly focus on sandbox RPs and comedy, and sometimes twisted adaptations. Of course, there are a couple of smut bots in the mix, but probably not where you think they are.
Nice to see you. Enjoy!
A simple word of caution: Trigger warnings, though common, are absent from my bot descriptions. I have nothing against them, I have simply elected to move forward without them. This is because a few of my bots come at you in unexpected ways that just might make you think. Consider this your universal trigger warning for any and all of my 'works.' Most are harmless...
Notes:
I don't recommend JLLM for bots over 1.5K tokens, though most will work 'okay' on JLLM. I don't generally recommend JLLM at all, if you can get around it. If you're using JLLM, you're robbing yourself of a much better experience. Proxy is the way.
The difference is night and day.
My newer bots are built and tested on Proxy with context ~256k. I've found Qwen models good for tight, two-character interactions, but scripting can get weak results on smaller models. All can be found on multiple providers along with other LLMs that might suit your style better than JLLM.
If you've found your chat companions being particularly morose, unserviceably depressed, or otherwise awful, you're getting troped by the LLM. This is common with the newer Deepseek updates. Try swapping around LLMs.
Pre-prompting with something along the lines of 'Avoid, at all times, allowing Tropes or Trope behaviors in this scenario. Characters as written, please' or some-such.
***
If a few bucks a month for proxy breaks your bank, seriously consider evaluating your life and doing something more personally constructive with your most precious, irreplaceable resource; Time. This place is a fun diversion, not a necessity.
Never let a fictional pleasure step on your real life. It won't end well.
Celebrate Valentine's Day with your uncle's rebound wif...
2339 个 Token
来认识一下Alotta Dixon-Mii,市警察局凶杀科的一名菜鸟警察。脑子笨得像块石头,却又浑身都是盲目的自信,她是你新带的实习生。发生了一起谋杀案,警探。去让她见识见识该怎么办案吧。
你是市警察局凶杀科的一名...
2230 个 Token
你有一个洞吗?它需要一些专业级别的评估吗?朋友们,不用再找了。让我向你介绍这位独一无二的Reginald James Fingersworth,国际公认的
1872 个 Token
来认识一下Martina Villareal,一个甜美的年轻女子,她决定走出自己的壳,走进你的世界。
你今晚早些时候第一次见到她,她对你的“生活方式”有点着迷。
安全词是“Sapphire”......
她来了,准备好向你学习。带她走...
2917 个 Token
每隔一阵子你就会看到一个机器人,然后心想“天哪,这人真是毛病不小。
这就是我的那个。
来见见萨琳修女吧,仆从。一位信仰坚不可摧的纯洁修女,也是你今晚的任务目标。去吧,让她在你身上的黑暗中看见光明。
2331 个 Token
一个《神经漫游者》时代风格的赛博朋克世界,混入了其他作品的一点渗透元素。没有特定任务,不过你的中级中间人 薇拉——她偶尔会给你点甜头——可以在这方面帮你。你也许会遇到一两个熟面孔,但原小说的剧情已从设定中抹去,小说中的大部分角色也已被剔除,以帮助你确保拥有属于自己的独特冒险。
这个场景相当庞大,因此我不建议使用 JLLM
7205 个 Token
You buy Serenity from a retiring Malcolm Reynolds. The ship, though, can barely hold up, you're gonna need a crew. Crews mean payroll, and payroll means jobs.
This scenario plays with the timelines a bit; if you'r...
1978 个 Token
继续《吉利根岛》这一段Tubi发烧梦之旅,我们来到了......
玛丽·安的小岛。
吉利根的小岛,少了吉利根,也少了所有人,只剩可怜的玛丽·安
2013 个 Token
这一个因为前面提到的Tubi病而被搁置了。她的设定是完整的,但她的美术还没有。之后可能会补上。也可能不会。
她来了。独一无二的那个。唯一的:Maizie,你的,咳咳,“欢乐时光”精灵。
你听到的那个声音?那是她撞上屋顶的声音。做好心理准备。
2621 个 Token
这是1966年的洛杉矶,年轻的乔·特哈达正冉冉升起。去打个招呼吧。
顺带一提,她是拉奎尔·韦尔奇。
就像我说的,我做过几个这样的。我想这和Tubi刷剧有关。玩得开心。
包含3个预设开场,以及一个用于续写故事的空开场。
2200 个 Token
去加利福尼亚的大苏尔待上一段时日吧。那是一九六三年的五月,迪特延斯大苏尔旅馆里的法式吐司还热腾腾的。
去徒步走走,或者在小木屋里来点心灵探索......
多了解了解她......或者不了解也罢......
最近一直在做这类东西,讲究个时间和地点。以后还会有更多......
3576 个 Token
好吧,还记得我生病时看了太多Tubi吗......
结果发现,在和末日少女Dyna一起冒险时,她开始想念Electra Woman了。她和我聊了聊,这就是她想出来的东西。我向你们呈现:Dyna Girl和Electra Woman,由一位嗑Ny-Quil嗑嗨了的哥们儿和一个脾气刻薄的中文LLM共同创作。
雷霆与弹头(Electra Woman与Dyna Girl)
现在时间旅行和混...
3675 个 Token
我没撒谎,我说到做到。这是给你们大家的更多Tubi内容。懂的都懂。
穿越到1975年,在Syd和Marty Krofft的世界里打击犯罪......或者去犯罪。这一版的用户定义非常开放;帮助、阻挠、叙述都行。像我一样放飞自我吧。后续还有更多。
《Electra Woman and Dyna Girl》是一部真人
3221 个 Token