The Evaluation Blind Spot: A
No mentions found
This entity hasn't been tracked yet, or Iris is still building its knowledge base.
Related Articles from SNS
The Evaluation Blind Spot: A Stereological Theory of Benchmark Coverage for Large Language Models
arXiv:2606.05169v1 Announce Type: new Abstract: We give a stereological theory of LLM benchmark coverage. For any suite with effective dimensionality d_eff, the visible Hausdorff distance between two convex capability profiles consistent with the same scores is bounded by epsilon + C R m^(-1/(d_eff-1)), with matching Lipschitz lower bound. Empirically, three independent leaderboards (Open LLM v2, an extended 12-benchmark suite, LiveBench) all have d_eff in [2.86, 4.80] on their competitive...
Trust, but Don't Verify: Epistemic Blind Spots in LLM Source Evaluation
arXiv:2606.05403v1 Announce Type: new Abstract: Language models increasingly act as epistemic proxies, synthesizing evidence from multiple sources to inform decisions. Whether they evaluate the quality of that evidence, or merely aggregate it based on surface presentation, remains poorly understood. We show that models possess the capability to detect fabricated statistics (correct identification rates of 0.76-1.00 for methodology in isolation) but do not recruit this capability during...
Biases in the Blind Spot: Detecting What LLMs Fail to Mention
Announce Type: replace Abstract: Large Language Models (LLMs) often provide chain-of-thought (CoT) reasoning traces that appear plausible, but may hide internal biases. We call these unverbalized biases. Monitoring models via their stated reasoning is therefore unreliable, and existing bias evaluations typically require predefined categories and hand-crafted datasets.
When Recovery Matters: The Blind Spot of Surrogate Privacy in MLLM Editing
arXiv:2606.07171v1 Announce Type: new Abstract: Multimodal Large Language Models (MLLMs) enable flexible instruction-driven image editing, but privacy risks arise when user images expose diverse and user-specific private content. Canonical privacy protection strategies typically substitute sensitive regions with surrogate content before cloud editing. Yet, the resulting output is often an edited surrogate rather than the desired edited source image, neglecting the local recovery in both...
Precision Is Not Faithfulness: Coverage-Aware Evaluation of Grounded Generation with a Complete Oracle
Announce Type: new Abstract: Reference-free faithfulness metrics verify each atomic claim a model makes against ground truth, and are increasingly used to evaluate grounded generation. We show they share a blind spot: they measure only precision -- are the stated claims supported? -- and therefore reward abstention, since a model can score near-perfect faithfulness by saying almost nothing.
Beyond Goodhart's Law: A Dynamic Benchmark for Evaluating Compliance in Multi-Agent Systems
Announce Type: new Abstract: The rapid evolution of Large Language Models (LLMs) from passive assistants to autonomous, execution-capable agents has introduced critical operational risks. Most current evaluation frameworks neglect procedural compliance, leading to ''Machiavellian'' behaviors where agents strategically violate safety rules to maximize rewards - a direct manifestation of Goodhart's Law. To address this blind spot, we introduce MAC-Bench, a dynamic, adversarial benchmark...
AdaGRPO: A Capability-Aware Adaptive Enhancement for Flow-based GRPO
Announce Type: new Abstract: Group Relative Policy Optimization (GRPO) has demonstrated remarkable success in aligning text-to-image (T2I) flow models with human preferences. However, we have identified that the learning loop of current flow-based GRPO is fundamentally decoupled from the learner's current capability, suffering from critical blind spots at both prompt selection and advantage estimation: (i) Existing methods sample prompts randomly, overlooking the substantial impact of data...
Deterministic Integrity Gates for LLM-Assisted Clinical Manuscript Preparation: An Auditable Biomedical Informatics Architecture
arXiv:2606.09500v1 Announce Type: new Abstract: Objective. Large language models (LLMs) increasingly draft clinical research manuscripts, but their fluency can hide fabricated citations, numbers that drift from source tables, and unmet reporting-guideline items. Existing tools generate text without verifying it, and self-critique inherits the blind spots that produce confident fabrication.
Send a SCOUT First: Pre-hoc Reasoning for Adaptive Detector Allocation in Prompt-Injection Defense
arXiv:2605.30837v1 Announce Type: new Abstract: Prompt-injection detectors are heterogeneous: each is strong on a different slice of attacks, and none is always reliable. Yet existing systems still treat detection as a fixed single-detector pipeline, committing every request to one detector's blind spots. We reframe defense as detector allocation: given a heterogeneous pool, decide per request which detectors to run and whether to escalate to an LLM judge.
Policy Improvement Reinforcement Learning
Announce Type: replace Abstract: Reinforcement Learning with Verifiable Rewards (RLVR) has become a central post-training paradigm for improving the reasoning capabilities of large language models. Yet existing methods share a common blind spot: they optimize policies based on instantaneous group-level or batch-level statistics without ever verifying whether the resulting update actually improved the model. This open-loop design -- updating in isolation at each step, guided only by...