朝花夕拾

Daily Vault

8月13日(木)（20件）

TRACE Bench: Task-driven Roleplay Agentic Checklist Evaluation

arXiv:2608.11236v1 Announce Type: new Abstract: Roleplay evaluation should do more than assign a single score: it should reveal which role requirements were tested, which failed, and which dialogue evidence supports the

·研究/·arxiv.org ↗

arxiv.org ↗

Retrofitting Recurrent Depth into a Pretrained Language Model: Installation, Extrapolation, Transfer, and Retention at Two Parameter Budgets

arXiv:2608.11233v1 Announce Type: new Abstract: A dense, pretrained language model can be retrofitted with recurrent depth and learn an iterative latent transition that persists after outcome-only annealing. Qwen2.5-0.5

·研究/·arxiv.org ↗

arxiv.org ↗

Backtrader-Bench: Benchmarking LLM Agents on Algorithmic Trading with Self-Generated MCQs

arXiv:2608.11232v1 Announce Type: new Abstract: Evaluating LLM coding agents in algorithmic trading is difficult because static benchmarks risk data contamination and numerical backtest outputs require ground truth from

·AI/, ·arxiv.org ↗

arxiv.org ↗

These mammal ancestors birthed live young far earlier than thought

Growth rings on a fossil bone suggest at least one mammal ancestor gave birth to live young rather than laying eggs 236 million years ago.

·研究/, ·sciencenews.org ↗

sciencenews.org ↗

Synchronizing Beliefs with Second-Order Theory-of-Mind in Human-Autonomy Teams

arXiv:2608.11229v1 Announce Type: new Abstract: Comparative feedback, asking people which of two behaviors they prefer, has become a standard way to align robot and agent behavior with human intent when the reward itsel

·研究/, ·arxiv.org ↗

arxiv.org ↗

Forecasting Side Effects of Activation Steering

arXiv:2608.11227v1 Announce Type: new Abstract: Activation steering modifies a language model by adding a learned direction to its hidden activations, enabling targeted behavioral changes without retraining. While effec

·研究/·arxiv.org ↗

arxiv.org ↗

Cutting AI Datacenter Energy with Reinforcement Learning: Measured Power Control of LLM Training from One GPU to the Fleet

arXiv:2608.11226v1 Announce Type: new Abstract: Reinforcement-learning post-training dominates modern language-model development, yet its power behavior on GPU hardware has not been characterized, and datacenters manage

·AI/, , , ·arxiv.org ↗

arxiv.org ↗

Identity from the Outside: A Conceptual Framework and Research Program for AI Personality Clones

arXiv:2608.11225v1 Announce Type: new Abstract: AI "personality clones" force a re-examination of personal identity in operational terms. Setting aside the hard problem of consciousness, we approach identity through the

·AI/·arxiv.org ↗

arxiv.org ↗

Harnessing agent memory to build lifelong AI partners for materials scientists

arXiv:2608.11224v1 Announce Type: new Abstract: Materials research advances through accumulated experience - scripts that work, protocols that are trusted, warnings attached to failed calculations or experiments, and ju

·AI/, , , ·arxiv.org ↗

arxiv.org ↗

A Conceptual Framework for Refining Influence Knowledge from Simulation Evidence in Cyber-Physical Systems

arXiv:2608.11221v1 Announce Type: new Abstract: Cyber-physical systems (CPS) are typically developed by multiple stakeholders who produce artefacts tailored to their specific domains of expertise. The behaviour of these

·国際/·arxiv.org ↗

arxiv.org ↗

LLMs in Process Diagram Engineering: From Optimal PFDs to Validated P&IDs

arXiv:2608.11220v1 Announce Type: new Abstract: Nowadays, the creation of a process flow diagram (PFD) and its subsequent transformation into a piping and instrumentation diagram (P&ID) is predominantly performed manual

·研究/, ·arxiv.org ↗

arxiv.org ↗

From Monolithic to Modular: Segment-level Automatic Prompt Optimization

arXiv:2608.11219v1 Announce Type: new Abstract: Automatic Prompt Optimization (APO) often rewrites prompts monolithically, which can improve one behavior while degrading others. We present SAPO, a segment-level APO meth

·研究/, ·arxiv.org ↗

arxiv.org ↗

MaSRead: Content-Addressed Reading of Replicated Latent Stores

arXiv:2608.11218v1 Announce Type: new Abstract: Independent agents that reason in latent space can share computed state as key-value cache fragments rather than text. Merged by a conflict-free replicated data type, thes

·研究/, ·arxiv.org ↗

arxiv.org ↗

AutoWorldModel-Bench: A State-Centric Benchmark for Automated World-Model Research

arXiv:2608.11216v1 Announce Type: new Abstract: World modeling is an unsettled field: architectures, training objectives, and state representations interact in complex ways, and no single recipe dominates across environ

·研究/·arxiv.org ↗

arxiv.org ↗

Poor Man's Agentic Modeling: Simulating Large LLM-Agent Societies on a Laptop

arXiv:2608.11215v1 Announce Type: new Abstract: Simulating societies of many large language model (LLM) agents is expensive, yet the questions asked of such simulations are usually macroscopic: phase behaviour, stylised

·AI/, , ·arxiv.org ↗

arxiv.org ↗

Detecting a Route Flip Is Easier Than Knowing Whether to Fix It: Causal Route-Mediated Damage in Quantized Mixture-of-Experts

arXiv:2608.11212v1 Announce Type: new Abstract: Top-k Mixture-of-Experts (MoE) routing is discontinuous, so a deployment-motivated numerical disturbance -- simulated 4-bit KV-cache quantization read by a protected BF16

·研究/·arxiv.org ↗

arxiv.org ↗

A Forced-Structure Reduction and Verifiable Bounds for Conway's 99-Graph

arXiv:2608.11211v1 Announce Type: new Abstract: Conway's 99-graph problem asks whether a strongly regular graph with parameters $(99,14,1,2)$ exists. We report a systematic, fully reproducible attack by an autonomous AI

·AI/·arxiv.org ↗

arxiv.org ↗

Distribird: Literature-Informed Prior Distribution Design for Bayesian Model Calibration

arXiv:2608.11210v1 Announce Type: new Abstract: Bayesian calibration of process-based models requires a prior distribution for each model parameter. Despite decades of methodological work, researchers almost always fall

·研究/·arxiv.org ↗

arxiv.org ↗

Dynamic Governance of Multi-LLM Agent Systems for Collaborative Conversational Outcomes

arXiv:2608.11207v1 Announce Type: new Abstract: When two LLM agents with structurally opposed objectives interact across multiple turns, the absence of a shared goal function produces not competition but collapse: the v

·AI/, , ·arxiv.org ↗

arxiv.org ↗

Limited Run Announces Lord Of The Rings: War In The North Physical Release

And so it begins. Earlier this week, Aspyr revived The Lord of the Rings: War in the North - Legacy Edition for Switch, Switch 2 and other platforms. Read the full article on nintendolife.com

·ACGN/·nintendolife.com ↗

nintendolife.com ↗