Aamer Mihaysi

O96a's activity

• All
• Models
• Datasets
• Spaces
• Buckets
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

commented a paper about 18 hours ago

FlowState: Execution State as Memory for Long-Horizon LLM Agents

Paper • 2609.34565 • Published 14 days ago • 16 •

4

commented a paper 2 days ago

Token-Set Choice Confounds POPE: A Systematic Audit of Yes/No Extraction in VLM Hallucination Evaluation

Paper • 2610.25964 • Published 3 days ago • 4 •

4

commented a paper 3 days ago

Embeddings Are Overpriced? A Frozen 16 MB Vocabulary for Small and Ternary Language Models

Paper • 2610.22428 • Published 7 days ago • 20 •

4

commented a paper 6 days ago

When the KV Cache Is Slower Than the GPU: Scaling Remote KV Tiers for Context-Parallel Prefill

Paper • 2610.21500 • Published 7 days ago • 4 •

3

commented a paper 8 days ago

EMHO: EMbodied Agent Harness Optimization via Experience Traces

Paper • 2610.08432 • Published 16 days ago • 11 •

4

commented a paper 10 days ago

When Do We Need On-Policy Distillation? Distilling on Offline Student Rollouts Is Often Better

Paper • 2610.11291 • Published 14 days ago • 28 •

3

commented a paper 11 days ago

Incidental information contaminates patient notes and disrupts clinical reasoning in large language models

Paper • 2610.08585 • Published 16 days ago • 7 •

3

commented a paper 12 days ago

Skill Constellations: Tracing the Supply Chain of Agent Skills on GitHub

Paper • 2610.11169 • Published 14 days ago • 8 •

3

commented a paper 13 days ago

Memento 3: Model-Based Recursive Self-Improvement through Reflective Rulebooks

Paper • 2610.11794 • Published 14 days ago • 48 •

5

commented a paper 14 days ago

A self-learning scientific agent for X-ray diffraction

Paper • 2610.07862 • Published 16 days ago • 19 •

6

commented a paper 15 days ago

Source Identification Is Not Fitness Testing: Measuring the Limits of Synthetic-Data Attribution

Paper • 2610.00417 • Published 22 days ago • 2 •

4

commented a paper 16 days ago

UndoBench: Separating Task Competence from Recovery Capability in Tool-Using AI Agents

Paper • 2610.05622 • Published 18 days ago • 22 •

3

commented a paper 17 days ago

VeriHarness: Scaling Agentic Verification for Long-Horizon Tasks

Paper • 2610.00972 • Published 21 days ago • 63 •

3

commented a paper 21 days ago

Safety of Latent Communication in Multi-Agent Systems

Paper • 2609.39788 • Published 22 days ago • 10 •

2

commented a paper 22 days ago

SEAD: A State-Based Perspective on Attack and Defense in Tool-Using Agents

Paper • 2609.34518 • Published 24 days ago • 4 •

3

commented a paper 23 days ago

Nereus: Adaptive Parallelism for LLM Post-Training

Paper • 2609.34645 • Published 24 days ago • 20 •

3

commented a paper 28 days ago

Six Layers Less: Encoder Pruning for Whisper with Label-Free Recovery

Paper • 2609.27980 • Published 29 days ago • 6 •

3

commented a paper 29 days ago

The Tasteful Agent: Measuring and Improving Taste in Long-Horizon Tasks

Paper • 2609.25804 • Published about 1 month ago • 165 •

6

commented a paper 30 days ago

Document Retrieval-Aware Chunking (D-RAC): Universal Retrieval-Aware Ingestion of Enterprise Documents via PDF Normalization and Multimodal Markdown Conversion

Paper • 2609.24220 • Published Sep 21 • 68 •

4

commented a paper about 1 month ago

Retention-Constrained Post-Training Quantization of Cellpose-SAM for Stem Cell Microscopy

Paper • 2609.21038 • Published Sep 17 • 5 •

4