ALT Multi-turn jailbreaks use feedback from a victim large language model to adapt their queries over several turns. However, once an effective trajectory has been found, it is unclear how much of the earlier query history is still needed. In this work, we study this by removing early attacker queries and running the remaining queries again in a new conversation while regenerating the victim's responses. We define the effective horizon as the shortest remaining query sequence that recovers the same attack progress. Empirically, we find that roughly half of the original query sequence is sufficient on average to recover the same attack progress. We call this phenomenon effective-horizon compression. We show that this shorter history can recover the best point reached during an unsuccessful attack and provide a better starting point for continuing the attack. It also helps identify which earlier queries should receive less learning signal during attacker training, and downweighting those que

ALT Large Language Models (LLMs) have demonstrated remarkable success in complex reasoning through explicit Chain-of-Thought (CoT), but the associated computational overhead and token verbosity remain significant barriers to efficiency. While implicit reasoning in continuous latent spaces offers a promising alternative, existing methods like Coconut and CODI often rely on fixed-length latent sequences and suffer from representational collapse. In this paper, we propose Sep-Latent, an adaptive implicit reasoning framework that utilizes separator tokens as structural semantic anchors. By introducing an Attention-Decay mechanism, we force the model to aggregate and compress reasoning logic into these anchors during training. We then align explicit and implicit pathways through multi-level distillation across all layers to ensure logical consistency. Crucially, Sep-Latent enables dynamic reasoning depth by adaptively terminating the latent thinking phase based on query complexity. Experimental

ALT Open-ended tool agents can complete the same objective through different tool paths, and their quality depends jointly on the interaction trace, environment state, and final artifacts, so such tasks rarely admit an executable ground-truth reward. The central challenge is therefore to construct a reward from existing experience that exposes the current policy's failures while remaining stationary during optimization. We introduce Reward Engine, which, like a self-motivated student, first works out what success requires and then optimizes toward it. From the task and initial trajectories, it builds a RewardPack of testcases for required correctness and rubrics for graded quality, each executed as a check that returns a judgment with evidence. The frozen RewardPack then guides the agent to revise and select its outputs, and can also serve as a reward for reinforcement learning. On WorkBuddy Bench Office and Claw-Eval, the constructed reward largely agrees with the benchmark graders, not r

ALT Long-horizon agents learn from failure, but the value of a failure record can change as execution progresses. We study when retaining, compressing, or removing the same history block improves subsequent execution relative to full-history KEEP. We introduce RECAP, a framework that combines matched context interventions, execution-relational prediction, and independent policy certification. Matched environment states and rollout seeds measure success and continuation prompt-cost differences; prefix-only relations between a failure block and the current execution propose promising edits. Independent task bundles then select a policy subject to harmful-intervention and non-positive-benefit risk constraints under a fixed measurement protocol. A five-seed longitudinal case holds both the failure block and its summary fixed: compression fails at an early state and becomes a cost-saving intervention at a later state. Across five backbones, RECAP produces positive paired-success saving estimate

ALT Parameter-Efficient Fine-Tuning (PEFT) has emerged as a prevalent strategy for adapting large pre-trained models. Among PEFT techniques, reparameterization-based methods, particularly those that employ low-rank or sparsity structures, have attracted significant attention. However, these approaches often underperform full fine-tuning. Specifically, low-rank methods constrain the update rank, which is misaligned with the inherently high-rank nature of full fine-tuning. In contrast, sparsity-based techniques limit the flexibility of the parameter space and compromise connectivity across weights. To overcome these limitations, we explore the use of Toeplitz matrices, whose entries are constant along each diagonal, therefore providing a compact parameterization without imposing explicit rank or sparsity constraints. A straightforward approach is to update weights via a product of Toeplitz matrices. However, similar to RNNs, long multiplicative chains often lead to gradient instability. To a

ALT Time series forecasting models are increasingly deployed in continuously evolving environments. After a forecast is issued, future observations gradually arrive and reveal how the model has failed. Existing methods mainly treat such observations as supervision for parameter updates. However, it remains unclear whether verified forecasting errors can themselves provide causal feedback for correcting future predictions. To address this, we introduce Error Feedback Calibration (EFC), a fully test-time framework for deterministic post-hoc correction of time series forecasting. EFC constructs a Causal Error Diagonal (CED) from revealed errors of overlapping historical forecasts, providing a fresh view of the frozen forecaster's recent failure behavior. Then Cumulative Error Transport (CET) maps this error state to future residuals using cumulative estimation, while temporal out-of-sample validation controls the correction strength. Temporal Residual Learning (TRL) further predicts only the

ALT Federated Prompt Tuning (FPT) adapts large vision–language models by freezing the pretrained backbone and optimizing only lightweight prompt parameters across clients. Although this design improves communication and parameter efficiency, it creates an overlooked security risk. Since the backbone is shared and frozen, a malicious client can induce backdoor behavior through the representation space while keeping its uploaded prompt updates close to benign ones. We propose ShadowFPT, a targeted backdoor attack that exploits this frozen-backbone attack surface. ShadowFPT first pretrains a learnable Shadow Trigger against the frozen CLIP visual encoder, using either auxiliary public data or the malicious client's local data, so that triggered inputs are steered toward the target class in representation space. During federated prompt tuning, the malicious client adapts the trigger under the current global prompt and then optimizes its local prompt on both clean and triggered samples. Only pr

ALT LLM agents are increasingly integrated into real-world works, yet they must handle diverse domain-specific tools and complex multi-sub-task problems that sequential tool invocation resolves inefficiently and with severe context bloat. The Main-Agent Delegation and Sub-Agent Execution (MAD-SAE) paradigm addresses these issues by delegating sub-tasks to specialized sub-agents for parallel execution in isolated contexts, but its delegation and orchestration quality remains unquantified. We introduce MS-Eval, a unified evaluation kit comprising an MCP-intensive benchmark, a controlled execution harness, tailored delegation and orchestration metrics, and the MSQ index for comprehensive assessment. Evaluating four MAD-SAE variants and eight frontier models reveals that (1) sub-agent specialization (Spec) is the best practical practice; (2) open-source models are largely competitive with few exceptions; and (3) MAD-SAE generally raises cost with contingent payoff, benefiting most open-source

ALT In branch-and-bound for Mixed-Integer Linear Programming (MILP), branching and node selection are two critical and tightly coupled decisions, yet most existing learning-based approaches focus on optimizing them separately. We propose Evo-B&B, an LLM-guided evolutionary framework that jointly evolves executable branching and node selection policies using end-to-end SCIP solving feedback. To improve search efficiency and reduce redundant evaluations of behaviorally similar policies, Evo-B&B introduces a behavior-aware multi-island evolutionary mechanism. It measures policy similarity from branching and node-selection behaviors on a fixed set of B&B decision states and uses this similarity to preserve population diversity. We evaluate Evo-B&B on four standard MILP benchmarks across multiple instance scales. Experimental results show that Evo-B&B achieves competitive solving performance and generalizes well to larger-scale instances, while ablation studies further demonstrate the effective

ALT World models have shown considerable promise in sequential decision-making by anticipating the consequences of actions and providing predictive foresight for policy learning. However, learning to reason about counterfactual actions remains challenging, since predictive objectives may underemphasize information critical for action evaluation, while unexecuted actions lack observed futures for direct supervision. To address these challenges, we propose the World Action Critic (WAC) Model, which proposes multimodal action candidates and jointly learns candidate-conditioned future prediction and consequence evaluation. In our proposed framework, critic supervision directly shapes intermediate predictive representations during rollout, encouraging them to retain information needed to distinguish action quality. Observed futures supervise prediction for executed actions, while external task-quality targets provide evaluation signals for counterfactual candidates. Through joint optimization,

ALT Emergent Misalignment - the out of domain generalization of narrow harmful fine-tuning first shown by Betley et al. [2025] has been widely studied in language models but not in reward models that supervise them. In this paper, we fine-tune two state-of-the-art reward models, Skywork-Reward-V2-Llama-3.1-8B and Skywork-Reward-V2-Qwen3-8B [Liu et al., 2025] and show that narrow fine-tuning reward models to prefer harm in one domain (medical or financial) generalizes to seven held out domains. This is not degeneracy, as the fine-tuned reward models lose only 4.8 and 5.3 points overall (RM-Llama) on the standard RewardBench 2 evaluation [Malik et al., 2025] while still outperforming the unmodified RM-Qwen model. We then fine-tune the underlying language models, Llama-3.1-8B-Instruct and Qwen3-8B using DPO [Rafailov et al., 2023] and SFT on the same data. We find that reward models are more vulnerable to preference inversions compared to their language model counterparts. We then ask the que

ALT Automated agent design increasingly trusts a single score to decide whether a change is an improvement – but a score optimized aggressively enough ceases to measure what it was meant to. When an agent also explains its design, that explanation is a second signal we can check without human labels: negate each stated claim, re-run, and observe whether performance drops. We ask whether such a verifier stays reliable once an agent optimizes against it. Our contribution is not a new verifier but a way to stress-test one: a pressure ratio ρ measuring how much the verifier steers the agent's search, and four falsifiable probes covering four failure modes – giving no signal, being too weak to matter, being gamed by stating fewer claims, and being queried before the agent commits. We apply the test to our own verifier, the Rationale Consistency Score (RCS), inside an LLM agent that designs machine-learning pipelines, with one generator model, six tasks, and three seeds. First, in our prior conf

ALT Retail recommenders rank relevance, but retailers earn contribution margin: an upsell earns the margin difference over the item it replaces, whereas a cross-sell earns its full margin. We introduce joint stochastic basket editing, a formulation in which one policy selects replacement and addition offers for an existing basket, values them by expected incremental contribution margin, and couples them through displayed-offer, accepted-offer, and joint-acceptance response channels. To our knowledge, it is the first formulation in which upsell and cross-sell responses are coupled; prior separable treatments are its zero-coupling special case. In real-data-calibrated simulations on five public basket datasets, StoBasket captures essentially all attainable offer margin: within 0.5% in every default-price setting and within 0.7% when the true response lies outside its model family. Correct economic accounting is the broadest gain: margin-weighted popularity leaves up to 11% of attainable marg

ALT Benchmarks drive AI progress, but are hampered by task under-specification and over-specification. Task under-specification causes guesswork, where solutions have to make things up, or worse, resort to contamination or reward hacking. Task over-specification causes spoilers, where solutions can bypass some of the harder steps the benchmark was supposed to measure. Since future-looking benchmarks should be difficult, over-specification reduces a benchmark's shelflife of community adoption. Repository-level coding benchmarks have helped the field progress but experience unnecessary churn due to both over- and under-specification. For instance, SWE-bench Pro, FEA-bench, Doc2Feat-bench, and Multi-SWE-bench all tackle under-specification by augmenting task statements with hints derived from ground truth. We show that leaking ground-truth information into the task statement causes over-specification in some of these benchmarks. We experiment with various specification levels, trying to threa

ALT In recent years, agentic orchestration has been widely applied to solve complex human tasks. However, existing orchestration learning still faces key challenges, including hand-designed curricula, advantage collapse when a task's rollouts all reach the same outcome, and unreliable evaluation signals that judge each skill by one noisy score. To address these challenges, we propose R²Orch, a new paradigm of Recursive Self-Play Orchestration: an inner loop of self-play that improves the orchestrator on self-chosen tasks, and an outer loop that improves how the inner loop improves. To support this paradigm, we introduce Recursive Learnability, a training objective that rewards a Proposer for choosing tasks whose rollouts keep reaching different outcomes as skills are withdrawn. Built on this objective, we further propose Skill Withdrawal, in which the outer loop withdraws each skill once a controlled evaluation within self-play shows that the orchestrator performs as well without it. Moreo

ALT The way a knowledge graph is serialized as text for inclusion into LLM context can substantially affect downstream LLM performance, with prior work reporting tens of percentage points variations across formats. However, existing studies consider only a limited set of representations and predate frontier reasoning models, which now offer stronger reasoning, instruction-following, and long-context capabilities. We evaluate 15 graph representation formats across nine frontier LLM configurations on ontology-based knowledge extraction and structural graph reasoning tasks. While knowledge extraction is relatively insensitive to different serialization formats, the choice of the optimal format for structural reasoning tasks significantly affects accuracy. We demonstrate that the optimal representation choice depends on model, query, and task and propose to approach graph serialization as a dynamic orchestration decision.

ALT Large language models exhibit misbehaviors that hinder deployment in high-stakes settings. One approach to monitor such failures is to use an external LLM as a monitor, but this is prohibitively expensive. A well-known and efficient class of monitors can detect such failures from internal representations (such as activation probes), but are tied to the model they were trained on, requiring costly re-annotation as new LLMs appear. A natural way to avoid this repeated annotation cost is to develop internal-representation based monitors that transfer zero-shot across LLMs. However, such transfer is challenging because internal representations are not naturally aligned across models; in particular, attention maps, known to contain signals of misbehavior, may differ in head function, ordering, and model depth. In this work, we characterize those cross-model attention mismatches as algebraic symmetries and use geometric deep learning principles to build LEMON (Language-model Equivariant MOni

ALT Text-based world models have recently shown potential to simulate complex RL environment state transitions for a diverse set of use cases. However, no standardized benchmark exists to compare, study and improve these models. Through this paper, we attempt to solve this problem by building GlobeBench, a novel benchmark with 1,632 tasks across five separate domains and nineteen simulation skills. The tasks check whether models can predict the environment's response to a specific action given the full agent trajectory by understanding earlier state changes, following environment rules, and predicting the effects of actions on both related and unrelated objects. We compare the predicted environment response with an execution-derived reference, using a hybrid architecture with both deterministic gates and a semantic judge. Among 20 general-purpose LLMs and specialized text world models, GPT-6 Astra achieves the highest task success rate at 49.57%. We find that model accuracy decreases as ac

ALT Multimodal large language models (MLLMs) offer scalable annotations of social and affective content, but their temporal agreement with human judgments and neural predictive value require further evaluation. We compared diffusion-based and autoregressive MLLMs with continuous human Theory of Mind (ToM) and affective valence ratings across three short cartoon movies. Behavioral analyses assessed temporally aligned Pearson correlations, rating distributions, and the effects of preceding scene descriptions. Ratings from each source were separately used in subject-specific regression models to predict regional fMRI responses. Model ratings captured aspects of human temporal variation despite systematic distributional differences, with generally stronger alignment for valence than ToM. Preceding context improved alignment in some conditions, depending on the model and movie. In The Present, more regions showed significant encoding for MLLM ratings than for human ratings, whereas effects were

ALT Verifier-guided self-repair conditions a model on feedback from failed attempts before it tries again, without weight updates. Yet two findings in prior work are in tension: verifier feedback can improve repair of failures, but self-repair itself can provide little-to-no advantage over independent resampling. We reconcile these findings in Lean theorem proving through controlled one-step experiments that hold the preceding failure fixed while varying whether the failed proof is shown, whether the compiler error corresponds to that failure, and the retry instruction. We test frozen Qwen3.6-27B, a general-purpose model, and DeepSeek-Prover-V2-7B, a specialized prover, on difficult MiniF2F problems. We find two opposing effects. Showing the model its failed predecessor reduces success and sharply increases copying of it, while on Qwen adding that attempt's compiler error recovers the loss, to the level of the diagnostic alone. Replacing the predecessor with another theorem's failed proof