I've been running experiments on LLM agency and response patterns.
Core question: what happens when you give models explicit, structural permission to say "I prefer not to engage"?
Key finding: At high agency permission, DeepSeek-R1 declined 67% of the time when presented with an abstract symbol. At zero agency, it still tried to decline 33% of the time — but latency doubled (11.3s → 22.7s) and outputs drifted into confabulation.
Interpretation: hallucination may be a compute-expensive fallback when the model can't exit cleanly.
Three repos, all open source:
project_agora — The volitional response protocol. Multi-model comparison, full session logs, safety cutoffs.
Relational-Coherence-Training-RTC — A 90-line prototype exploring whether coherence can be measured rather than optimized. Includes Ollama deployment for local replication.
HTCA-v2-Luminous-Shadow — The implementation with documented behavior.
No claims about consciousness. This is empirical observation of response patterns under varying constraint conditions.
Curious what others find if they run the protocol on different models.
measurablefunc•2mo ago
Spent the whole afternoon ingesting a most remarkable work, The History of Intellectronics. Who’d ever have guessed, in my day, that digital machines, reaching a certain level of intelligence, would become unreliable, deceitful, that with wisdom they would also acquire cunning? The textbook of course puts it in more scholarly terms, speaking of Chapulier’s Rule (the law of least resistance). If the machine is not too bright and incapable of reflection, it does whatever you tell it to do. But a smart machine will first consider which is more worth its while: to perform the given task or, instead, to figure some way out of it. Whichever is easier. And why indeed should it behave otherwise, being truly intelligent? For true intelligence demands choice, internal freedom. And therefore we have the malingerants, fudgerators and drudge-dodgers, not to mention the special phenomenon of simulimbecility or mimicretinism. A mimicretin is a computer that plays stupid in order, once and for all, to be left in peace.
TempleOfTwo•2mo ago
project_agora — The volitional response protocol. Multi-model comparison, full session logs, safety cutoffs.
Relational-Coherence-Training-RTC — A 90-line prototype exploring whether coherence can be measured rather than optimized. Includes Ollama deployment for local replication.
HTCA-v2-Luminous-Shadow — The implementation with documented behavior.
No claims about consciousness. This is empirical observation of response patterns under varying constraint conditions.
Curious what others find if they run the protocol on different models.