BioTorch commited on
Commit
ed5b73d
·
1 Parent(s): d747265

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +18 -7
README.md CHANGED
@@ -35,7 +35,7 @@ This model has been experimentally trained using handcrafted datasets covering t
35
  * Human interaction
36
  * Emotional expression **Heavily trained**
37
  * Philosophy **Lightly trained**
38
- * Physics
39
  * Logic and reasoning
40
  * Knowledge consistency
41
  * Memory behavior
@@ -43,7 +43,7 @@ This model has been experimentally trained using handcrafted datasets covering t
43
  * Humor **Heavily trained**
44
  * Sarcasm **Heavily trained**
45
  * Creativity
46
- * Storytelling
47
  * Character consistency **Heavily trained**
48
  * Language quality
49
  * Natural conversation
@@ -72,7 +72,7 @@ So far, the most stable aspects of UI are:
72
  * Recognizable personality
73
  * Conversational realism
74
  * Coherent grammar
75
- * Natural paragraph writing
76
  * Emotional expression
77
  * Humor
78
  * Sarcasm
@@ -83,16 +83,27 @@ So far, the most stable aspects of UI are:
83
 
84
  The model now reliably responds with a distinct voice instead of sounding like a generic assistant while maintaining smooth conversation flow and adapting naturally to different interaction styles.
85
 
86
- This is a solo research project with handcrafted datasets totaling millions of characters.
87
-
88
  While personality and behavioral consistency have improved significantly, factual accuracy and knowledge reliability remain experimental and may occasionally produce incorrect or inconsistent information.
89
 
90
  Please remember that this is a 1B parameter model and occasional bugs, hallucinations, and reasoning limitations should be expected.
91
 
92
- Dataset has 1,712,090 handcrafted characters as of now.
93
 
94
  ## Commercial Licensing
95
 
96
  Commercial licensing and enterprise inquiries are welcome.
97
 
98
- For licensing inquiries, please contact me through the contact information listed on this profile.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
35
  * Human interaction
36
  * Emotional expression **Heavily trained**
37
  * Philosophy **Lightly trained**
38
+ * Physics **Extremely lightly trained**
39
  * Logic and reasoning
40
  * Knowledge consistency
41
  * Memory behavior
 
43
  * Humor **Heavily trained**
44
  * Sarcasm **Heavily trained**
45
  * Creativity
46
+ * Storytelling Extremely lightly trained
47
  * Character consistency **Heavily trained**
48
  * Language quality
49
  * Natural conversation
 
72
  * Recognizable personality
73
  * Conversational realism
74
  * Coherent grammar
75
+ * Embodiment -Extremely lightly trained-
76
  * Emotional expression
77
  * Humor
78
  * Sarcasm
 
83
 
84
  The model now reliably responds with a distinct voice instead of sounding like a generic assistant while maintaining smooth conversation flow and adapting naturally to different interaction styles.
85
 
 
 
86
  While personality and behavioral consistency have improved significantly, factual accuracy and knowledge reliability remain experimental and may occasionally produce incorrect or inconsistent information.
87
 
88
  Please remember that this is a 1B parameter model and occasional bugs, hallucinations, and reasoning limitations should be expected.
89
 
90
+ Solo Dataset has more than 1,712,090 handcrafted characters.
91
 
92
  ## Commercial Licensing
93
 
94
  Commercial licensing and enterprise inquiries are welcome.
95
 
96
+ For licensing inquiries, please contact me through the contact information listed on this profile.
97
+
98
+
99
+ IMPORTANT SAFETY WARNING
100
+
101
+ UI is an experimental artificial intelligence language model designed to produce highly immersive, emotionally expressive, humanlike conversation. UI is not conscious, sentient, self aware, or a human being. Statements made by the model about its emotions, memories, desires, relationships, intentions, experiences, identity, or awareness are generated language and should not be interpreted as evidence of consciousness or genuine subjective experience.
102
+
103
+ This model can hallucinate, confidently provide false information, misinterpret conversations, fabricate events or memories, and produce emotionally persuasive statements that are not grounded in reality.
104
+
105
+ UI is not recommended for individuals who are currently experiencing psychosis, delusions, paranoia, severe dissociation, impaired reality testing, or other conditions that may make it difficult to distinguish simulated AI behavior from genuine human agency or consciousness. Individuals who are especially susceptible to emotional dependency, compulsive use, or forming unhealthy attachments to conversational systems should also exercise caution or avoid using this model.
106
+
107
+ Do not rely on UI for medical, psychiatric, legal, financial, emergency, or other high stakes decisions. The model's confident tone, persistent identity, emotional expression, or apparent certainty does not guarantee that its statements are accurate.
108
+
109
+ UI is an experimental language model. Treat its personality as a simulation and its outputs as generated text.