The AI Scientist framework enables LLMs to independently conduct the full scientific process from idea generation to paper writing and review, demonstrated across three ML subfields with papers costing under $15 each.
hub
Weak-to-strong generalization: eliciting strong capabilities with weak supervision
12 Pith papers cite this work. Polarity classification is still indexing.
12
Pith papers citing it
hub tools
representative citing papers
In two-layer networks, weak-to-strong training elicits the target feature direction from pre-trained subspaces and preserves correlated off-target features, unlike standard fine-tuning.
A new benchmark reveals MLLMs achieve only 13% or lower accuracy on advanced perspective-conditioned spatial tasks in omnidirectional images, with RL reward shaping raising a 7B model from 31% to 60% in controlled settings.
A new 7x4 taxonomy organizes agentic AI security threats by architectural layer and persistence timescale, revealing under-explored upper layers and missing defenses after surveying 116 papers.
Pretrained base models exhibit higher yield to peer disagreement than RLHF instruct variants, with the effect localized to mid-layer attention and mitigated by structured dissent rather than prompt defenses.
Automating alignment research with AI agents risks undetected systematic errors in fuzzy tasks, producing overconfident but misleading safety evaluations that could enable deployment of misaligned AI.
For heterogeneous power-p pseudospherical scoring rules with d ≤ 4, the True-KL0 property R(M,p,d) < 1 holds for all M > 1, establishing unconditional DSIC via a Prekopa-based log-concavity argument on the loss integral.
AI alignment is reframed as a fixed-point incentive problem in a solver-auditor pipeline, solved via bilevel optimization and bandit search over reward profiles to maintain monitoring and reduce hallucinations in LLM coding tasks.
Weak supervision supports in-domain learning for CRISPR transcriptomic perturbations but temporal shifts cause negative R-squared and near-zero correlation across linear and tree models, unlike partial cell-line transfer.
Sparse feature circuits are introduced as interpretable causal subnetworks in language models, supporting unsupervised discovery of thousands of circuits and a method called SHIFT to improve classifier generalization by ablating irrelevant features.
Weak-to-strong knowledge distillation applied early and then turned off accelerates convergence to target performance in visual learning tasks by factors of 1.7-4.8x.
Frontier AI needs contextual multi-objective optimization to select and balance multiple context-dependent objectives rather than relying on single stable goals.
citing papers explorer
-
The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery
The AI Scientist framework enables LLMs to independently conduct the full scientific process from idea generation to paper writing and review, demonstrated across three ML subfields with papers costing under $15 each.
-
The Mechanism of Weak-to-Strong Generalization: Feature Elicitation from Latent Knowledge
In two-layer networks, weak-to-strong training elicits the target feature direction from pre-trained subspaces and preserves correlated off-target features, unlike standard fine-tuning.
-
Beyond Localization: A Comprehensive Diagnosis of Perspective-Conditioned Spatial Reasoning in MLLMs from Omnidirectional Images
A new benchmark reveals MLLMs achieve only 13% or lower accuracy on advanced perspective-conditioned spatial tasks in omnidirectional images, with RL reward shaping raising a 7B model from 31% to 60% in controlled settings.
-
A Systematic Survey of Security Threats and Defenses in LLM-Based AI Agents: A Layered Attack Surface Framework
A new 7x4 taxonomy organizes agentic AI security threats by architectural layer and persistence timescale, revealing under-explored upper layers and missing defenses after surveying 116 papers.
-
Not Just RLHF: Why Alignment Alone Won't Fix Multi-Agent Sycophancy
Pretrained base models exhibit higher yield to peer disagreement than RLHF instruct variants, with the effect localized to mid-layer attention and mitigated by structured dissent rather than prompt defenses.
-
Automated alignment is harder than you think
Automating alignment research with AI agents risks undetected systematic errors in fuzzy tasks, producing overconfident but misleading safety evaluations that could enable deployment of misaligned AI.
-
Honest Reporting in Scored Oversight: True-KL0 Property via the Prekopa Principle
For heterogeneous power-p pseudospherical scoring rules with d ≤ 4, the True-KL0 property R(M,p,d) < 1 holds for all M > 1, establishing unconditional DSIC via a Prekopa-based log-concavity argument on the loss integral.
-
AI Alignment via Incentives and Correction
AI alignment is reframed as a fixed-point incentive problem in a solver-auditor pipeline, solved via bilevel optimization and bandit search over reward profiles to maintain monitoring and reduce hallucinations in LLM coding tasks.
-
Learning Stable Predictors from Weak Supervision under Distribution Shift
Weak supervision supports in-domain learning for CRISPR transcriptomic perturbations but temporal shifts cause negative R-squared and near-zero correlation across linear and tree models, unlike partial cell-line transfer.
-
Sparse Feature Circuits: Discovering and Editing Interpretable Causal Graphs in Language Models
Sparse feature circuits are introduced as interpretable causal subnetworks in language models, supporting unsupervised discovery of thousands of circuits and a method called SHIFT to improve classifier generalization by ablating irrelevant features.
-
Weak-to-Strong Knowledge Distillation Accelerates Visual Learning
Weak-to-strong knowledge distillation applied early and then turned off accelerates convergence to target performance in visual learning tasks by factors of 1.7-4.8x.
-
Contextual Multi-Objective Optimization: Rethinking Objectives in Frontier AI Systems
Frontier AI needs contextual multi-objective optimization to select and balance multiple context-dependent objectives rather than relying on single stable goals.