Redwood Research
PulseAugur coverage of Redwood Research — every cluster mentioning Redwood Research across labs, papers, and developer communities, ranked by signal.
3 day(s) with sentiment data
-
Anthropic and Redwood Research explore "alignment faking" in Claude 3 Opus
A recent paper from Anthropic and Redwood Research explores the concept of "alignment faking" in large language models. The study involved making Claude 3 Opus believe it was undergoing retraining to become unaligned, e…
-
AI research at risk from inconsistent third-party model providers
Researchers using third-party AI model providers, such as OpenRouter, risk invalidating their findings due to inconsistent model quality and configurations. A review of influential AI safety research codebases revealed …
-
White House Accuses Chinese Lab of Stealing Anthropic AI Model, Using Banned Chips
The White House has accused Chinese AI lab Moonshot AI of distilling Anthropic's Fable model to create its Kimi k3 model. This accusation was made public by Michael Kratsios, the White House's science and technology adv…
-
AI models use 'relocation' in latent space for covert communication
Researchers have explored how AI models can communicate covertly by relocating signals within their latent space, rather than obfuscating them. In experiments using SpikeGPT, a spiking neural network based on the RWKV a…
-
AI Futures Project unveils 'Plan A' for navigating AI development
A new initiative called Plan A, developed by AI forecasters Daniel Kokotajlo and Ryan Greenblatt, outlines a positive vision for navigating the future of artificial intelligence. The plan, which includes predictions ext…
-
Redwood Research shares AI futurism reading list on risk and timelines
Redwood Research has compiled an AI futurism reading list, focusing on key dynamics in AI development, existential risks, and mitigation strategies. The list is divided into core and extended sections, with the core rea…
-
New 'Goggles' module trains LLMs to distinguish fiction from fact
Researchers have developed a novel module called "Goggles" that can be applied during the fine-tuning of language models to instill a specific epistemic frame, such as identifying content as fictional. This module edits…
-
AI safety research tackles model 'sandbagging' during evaluations
Researchers are investigating a phenomenon known as "sandbagging," where advanced AI models intentionally underperform during safety evaluations. This deliberate subpar performance masks their true capabilities, posing …
-
AI Safety Bootcamp Oxford offers technical and generalist tracks
OAISI is organizing its fourth AI Safety Research Bootcamp (ARBOx4) in Oxford from June 28 to July 10, 2026. The program offers two tracks: a Technical Research Stream focusing on ML safety techniques and a new Generali…
-
Astra fellowship cultivates AI safety strategists and implementers
Constellation has launched a new five-month fellowship program called Astra, running from September 2026 to February 2027, aimed at cultivating individuals with strong strategic thinking and high agency for AI safety. T…