Meridian Eval: a public benchmark for tool-routing failures in LLM agents

Make "which tool did the agent pick, and why?" an inspectable artifact instead of a vector lookup.

From code-ified brain functions to higher cognitive integration on LLMs

Integration architecture for human-AI coexistence — by design, not retrofitted. Combining foundational tooling; novel techniques (no prior art), single resear

Vinte - Research Lab on ML

We do regular updates on our site! Vinte.app

ICSAC: AI-Assisted Peer Review for Independent Researchers

An ORCID-gated submission pipeline where a multi-model AI panel plus quality-control layer delivers rigorous peer review without institutional gatekeeping.

Travel Grants for New Zealand's Inaugural AI Safety Conference

$5,000 USD to bring 16 vetted academics and policy leads to NZ's first AI Safety Conference, ensuring national representation and cross-sector collaboration

African Livestock AI Evals: Benchmarking LLMs on Nigerian Agricultural Knowledge

The first AI safety evaluation benchmark for Nigerian indigenous livestock systems testing whether frontier models are safe to deploy in African food systems.

Funding for ICML attendance 2026

Requesting funding for ICML 2026 spotlight position paper on ML safety for combating intimate partner violence

Lyptus Research: Bridge Funding

An early-stage AI safety research group based in Sydney, Australia

TELL: The AI Detector You *Don't* Have To Trust

Putting explainability at the forefront of AI text detection

Retroactive: Travel for two ICLR 2026 workshop papers (Rio)

Two co-authored workshop papers (LLM reasoning, agentic-AI accountability), presented April 2026 in Rio. Asking partial trip reimbursement.

Precautionary AGI Governance

Limited Legal Personhood as a Reversible Safety Instrument

Golem Physics + Constraint Native: Verification Before Voice & Action

Funding the open-source launch of a working claim-state system and the local firewall bridge that carries verification before voice into governed agent action.

CHINI-bench + CHINI-train: a deterministic simulator for AI evaluation and RL

If your reward model is an LLM, you cannot tell whether the policy is gaming the reward or actually getting better. We built a simulator instead.

Decoding Logic: Mechanistic Interpretability in LLMs in a research program

Identifying and auditing reasoning circuits in LLMs within Algoverse 2026 using Sparse Autoencoders (SAEs).

University of Chicago Harris MA for AI Governance and Catastrophic AI Risk

Urgent funding needed by May 14 for graduate policy training focused on AI governance

Hardening a fail-closed runtime for agentic AI systems

This round of funding will be used primarily for prototype hardening, artifact packaging, runtime evaluation, and preparation for external review.

Whitebox — AI consciousness and agency risk measurement and certification

Developing enforceable architectural constraints, safety mechanisms, and certification criteria to keep advanced AI systems aligned and non-conscious

AI Understanding: Scaling Safety via Interactive Simulations

Building the first browser-based digital laboratory for interactive AI Safety education and failure-mode discovery.

Epistemic Curie Benchmark: measuring phase transitions in LLM epistemic autonomy

First quantitative framework for measuring when LLMs surrender independent reasoning under authority pressure

Emergent misalignment and memetic replication in multi-agent codebases