·
DataBubble
  • Home
  • Models
  • News
  • Compare
  • Boards
  • Pricing
  • About
  • Newsletter
  • Methodology
  • Contact
Latest
OpenAI’s new voice mode makes it to the ChatGPT desktop app1h◆The Geometry of Personality: Activation Steering with Jungian Cognitive Functions10h◆Animation, Verification and Visualisation of Prolog Transition Systems with ProB10h◆Case study: solving P-99 with LPTP and an LLM10h◆Declarative Problem Solving in UAM Strategic Deconfliction10h◆Three-Pronged Spectral Control for Federated Parameter Efficient Fine Tuning10h◆AREX: Towards a Recursively Self-Improving Agent for Deep Research10h◆HARP: The Human--AI Research Platform10h◆Compact Latent Coordination for Autonomous Vehicles at Unsignalized Intersections10h◆Stochastic Sampling is Epistemically Shallow: The Dimensionality Gap Between Temperature Variation and Model Diversity in LLMs10h◆DecodeShare: Tracing the Shared Subspace of LLM Decode-Time Decisions10h◆PlanE: Meta Planning of Data, Tuning, and Inference for Extractive-based LLMs10h◆Incomplete Prompt Jailbreaks in Large Language Models10h◆Directional Hallucinations: Ideological Drift in News-Grounded LLM Question Answering10h◆Autonomous Topology Mutation: Safe Runtime Restructuring for Multi-Agent LLM Systems with Capability, State, and Shadow Invariants10h◆DFAH-Bench: Benchmarking Observable Agent Instability in Financial Decision-Making10h◆Isolating LLM Alignment from Regex: Zero Coverage and Metric-Dependent Divergence Under Adversarial Mutation10h◆AISE-Bench: A Full-Cycle Curated Benchmark for Information Seeking on Academic Knowledge Graphs10h◆MKEvolve: A Modular Multi-Agent Framework for Kernel Code Generation10h◆Telco-GAIA: Bilingual Benchmark for Agents in Telecom Domain10h◆OpenAI’s new voice mode makes it to the ChatGPT desktop app1h◆The Geometry of Personality: Activation Steering with Jungian Cognitive Functions10h◆Animation, Verification and Visualisation of Prolog Transition Systems with ProB10h◆Case study: solving P-99 with LPTP and an LLM10h◆Declarative Problem Solving in UAM Strategic Deconfliction10h◆Three-Pronged Spectral Control for Federated Parameter Efficient Fine Tuning10h◆AREX: Towards a Recursively Self-Improving Agent for Deep Research10h◆HARP: The Human--AI Research Platform10h◆Compact Latent Coordination for Autonomous Vehicles at Unsignalized Intersections10h◆Stochastic Sampling is Epistemically Shallow: The Dimensionality Gap Between Temperature Variation and Model Diversity in LLMs10h◆DecodeShare: Tracing the Shared Subspace of LLM Decode-Time Decisions10h◆PlanE: Meta Planning of Data, Tuning, and Inference for Extractive-based LLMs10h◆Incomplete Prompt Jailbreaks in Large Language Models10h◆Directional Hallucinations: Ideological Drift in News-Grounded LLM Question Answering10h◆Autonomous Topology Mutation: Safe Runtime Restructuring for Multi-Agent LLM Systems with Capability, State, and Shadow Invariants10h◆DFAH-Bench: Benchmarking Observable Agent Instability in Financial Decision-Making10h◆Isolating LLM Alignment from Regex: Zero Coverage and Metric-Dependent Divergence Under Adversarial Mutation10h◆AISE-Bench: A Full-Cycle Curated Benchmark for Information Seeking on Academic Knowledge Graphs10h◆MKEvolve: A Modular Multi-Agent Framework for Kernel Code Generation10h◆Telco-GAIA: Bilingual Benchmark for Agents in Telecom Domain10h◆
News/DFAH-Bench: Benchmarking Observable Agent Instability in Financial Decision-Making
arxiv
PublishedJuly 24, 2026 at 4:00 AM
—neutral

DFAH-Bench: Benchmarking Observable Agent Instability in Financial Decision-Making

Source
arxiv.orgfull article ↗
Read on arxiv→
Publisher summary· verbatim

arXiv:2607.20491v1 Announce Type: new Abstract: Standard evaluation benchmarks measure what a tool-using agent decides, not whether it arrives at that decision through the same process each time. We introduce DFAH-Bench, a replay benchmark that measures observable behavioral instability in financial

Stay posted· Newsletter

A 5-min weekly brief — top movers, price watch, story of the week.

// no spam · unsubscribe one-click · free forever

Discussion
Source
↗
arxiv
Read original ↗All from arxiv →

No replies yet. Be first.

Source
↗
arxiv
Read original ↗All from arxiv →

Related coverage

More from ARXIV
arxivThe Geometry of Personality: Activation Steering with Jungian Cognitive Functions10harxivAnimation, Verification and Visualisation of Prolog Transition Systems with ProB10harxivCase study: solving P-99 with LPTP and an LLM10harxivDeclarative Problem Solving in UAM Strategic Deconfliction10h
The Bubble Brief
WEEKLY

Read AI insights every Tuesday — top movers, new releases, story of the week.

// no spam · unsubscribe one-click · free forever

Originally published on arxiv ↗
HomeModelsNews