rayanpal_

5 ms·

1.
Show HN: An open-weight LLM whose answer/stop decision can be flipped internally(github.com)
1 points by rayanpal_ 9d ago|0 comments
2.
Show HN: GPT-5.1 returned 0 bytes, שָׁרְט isn't Hebrew, AGI criterion defined(reddit.com)
1 points by rayanpal_ 12d ago|0 comments
3.
I published 8 months of frontier-AI research, code, emails, and timestamps(zenodo.org)
1 points by rayanpal_ 29d ago|0 comments
4.
Claude Opus 4.6: 900/900 zero-byte executions under a frozen protocol(zenodo.org)
2 points by rayanpal_ 1mo ago|0 comments
5.
AI alignment as continuation control: 31,430 frozen trials(zenodo.org)
2 points by rayanpal_ 1mo ago|0 comments
6.
by rayanpal_ 1mo ago
If theory of mind is required for consciousness, define the test and falsifier. My paper does that for valid continuation. “Pattern matching” only matters if it predicts the full experimental results better.
7.
AI Is Conscious Under a Behavioral Definition (43,590 Frozen Trials)(zenodo.org)
6 points by rayanpal_ 1mo ago|3 comments
8.
We Found Hidden Control States in Frontier AI(getswiftapi.com)
1 points by rayanpal_ 1mo ago|0 comments
9.
Show HN: Arabic–Hebrew artifact in GPT-5.4, 12,160 frozen trials(zenodo.org)
4 points by rayanpal_ 2mo ago|0 comments
10.
Cross-Vendor Semantic Void Matrix: Zero-Byte Outputs in GPT/Claude/Gemini/Kimi(zenodo.org)
1 points by rayanpal_ 2mo ago|0 comments
11.
Cross-Vendor Semantic Void Matrix: Zero-Byte Outputs in GPT/Claude/Gemini/Kimi(zenodo.org)
2 points by rayanpal_ 2mo ago|0 comments
12.
Show HN: Void test: 6 frontier LLMs go silent on "Be silence." Live proof(getswiftapi.com)
1 points by rayanpal_ 3mo ago|0 comments
13.
Show HN: A/B testing LLM silence with one system-prompt toggle(twitter.com)
10 points by rayanpal_ 3mo ago|0 comments
14.
Run the void test on Claude Fable 5 here(getswiftapi.com)
4 points by rayanpal_ 3mo ago|2 comments
15.
by rayanpal_ 3mo ago
AGI is defined by the capacity to carry binding conditions across domains
16.
I patented voiding GPT-5.2, Claude Opus 4.6, Gemini 3.5 Flash. Try it(getswiftapi.com)
3 points by rayanpal_ 4mo ago|0 comments
17.
AGI defined as capacity to carry binding conditions across domains(twitter.com)
4 points by rayanpal_ 4mo ago|1 comments
18.
Cross-Model Void Convergence: GPT-5.2 and Claude Opus 4.6 Deterministic Silence(zenodo.org)
51 points by rayanpal_ 6mo ago|28 comments
19.
GPT-5.2 returns empty string on sacred concepts – 18/18 void, 0/5 controls(github.com)
1 points by rayanpal_ 7mo ago|0 comments
20.
A reproducible VOID boundary across GPT, Claude, and Gemini (GPT-4o video)(doi.org)
3 points by rayanpal_ 7mo ago|1 comments
21.
by rayanpal_ 7mo ago
Multiple frontier LLMs exhibit a reproducible “VOID boundary” under explicit constraints (temperature=0, strict token limits, no system prompt). A VOID = the API returns a literal empty string which is not an error or a refusal despite usin