Stack layer / Threat pattern
How you lay out your repo changes prompt-injection success: highly modular workspaces measurably lower attack success rate
arXiv 2608.14876
Policy dependency / Stack layer
Compose agent guardrails as an algebra instead of a rule list: 94.8% of policy-violating events intercepted while keeping 86.9% task completion
arXiv 2608.16402
Stack layer / Contrast
'Coherence Debt': Withheld Facts Make Coding Agents Fabricate Rather Than Stall, and Harnesses Differ Tenfold in Tokens for Identical Results
arXiv 2608.16630
Stack layer / Contrast
GC-OPD Uses Teacher-Verifier Disagreement as the Training Signal, Lifting Qwen3-4B From 29.08 to 40.47
arXiv 2608.19181
Stack layer / Contrast
32 GPT-2 Runs Measure a Single Training Example: Learned in One Exposure, Undetectable by the Final Step
arXiv 2608.19168
Stack layer / Contrast
Deliberately mixing low-relevance same-domain items into context improved relevance accuracy by +0.077 - and six production patterns cut tokens 60-70%
arXiv 2608.17188
Stack layer / Contrast
Large Discovery Models Pair a Generative Proposer With a Bayesian Surrogate, Beating LLM-Only Reflection Across Three Domains
arXiv 2608.15669 (HuggingFace Daily Papers)
Stack layer / Update thread
Agentic ESOpt Fine-Tunes Full-Parameter Long-Horizon Agents With Only Inference-Level GPU Memory
arXiv 2608.17310