← Back to all news
arXiv Computation and Language

Intelligence Under Time Constraints: Rethinking Test-Time Compute

The paper "Intelligence Under Time Constraints: Rethinking Test-Time Compute" examines how to decide not only how much computation to perform but also when to start it in streaming interactions where evidence arrives incrementally and may change. It introduces the information‑slack dilemma, describing the trade‑off between early computation that has more time to finish but relies on incomplete evidence, and waiting for more information that reduces computational slack. The authors propose a research agenda focused on selective recovery under controlled evidence revisions, and suggest evaluation metrics that separate early‑execution effects, deployment value versus full‑input alternatives, and the added value of predictive policies while accounting for shared‑resource costs, aiming for trustworthy, on‑time responses within a declared resource envelope.

Related stories

arXiv Computation and Language
Sep 15

Beyond Depth and Width: The Information-Slack Dilemma in Streaming Test-Time Compute

The paper discusses how the same computational task can require different reasoning strategies depending on the order in which evidence arrives, introducing the concept of an "information‑slack dilemma." It argues that early computation may be useful only if its benefits outweigh the costs of later verification, invalidation, and recovery, and proposes a research agenda focused on selective recovery and predictive policies. The authors emphasize evaluating these approaches by separating early‑execution effects, deployment value versus full‑input alternatives, and the added value of predictive policies while considering shared‑resource costs.

By Xiaotian Zhang (Trooly.AI)
More like this →
arXiv AI
6d ago

Scaling Lean Theorem Proving for Real-World Code Verification

The paper introduces EasyProver, a Lean theorem prover that uses recursive agents, isolated contexts, asynchronous search, parallel tactic probing, and an elastic elaboration service. It achieves near‑saturation results on several benchmark suites, proving 669/672 PutnamBench problems, all six IMO 2026 formalizations, 597/600 NTP4VC verification conditions, and all 189 VERINA tasks. Additionally, the authors present Aegis, a pipeline that transforms competitive programming problems into harder verification tasks by generating safe Rust solutions, translating them to Lean, and producing 2,045 Rust‑Lean program pairs with 11,533 candidate specifications, enabling new benchmarks for provers and reinforcement learning.

By Debargha Ganguly, Divya Raghunathan, Dejan Jovanovi\'c, Vikash Singh, Soonho Kong, Aaron Tomb, Aman Goel, Ferhat Erata, Sam Bayless
agentsreinforcement-learningbenchmarks
More like this →
arXiv Machine Learning
Sep 17

ASPIRE: Asynchronous Batched Self-Speculative Decoding for Long-Context LLM Inference

ASPIRE introduces a non‑synchronized batched self‑speculative decoding framework for long‑context LLM inference, addressing the memory bottleneck of attention by drafting tokens with sparse attention and verifying them with full attention. It combines a unified mixed forward pass, a lightweight online speculation scheduler that lets each request independently decide when to verify, and an intra‑draft refresh layer that updates the sparse context at every draft step. Experiments on three models and five benchmarks show 1.70–4.58× speedup over autoregressive baselines and a 27% average improvement over the strongest prior self‑speculative methods.

By Amir Ziashahabi, Hossein Entezari Zarch, Lei Gao, Murali Annavaram, Salman Avestimehr
llmsefficiencybenchmarks
More like this →
Hugging Face Trending Papers
Oct 29

Efficient Scalable Oversight via Rewinding

The paper introduces Rewindable Interactive Proofs (RIPs), a new model for scalable oversight where a weak verifier can rewind a powerful AI prover to an earlier state during interaction. RIPs enable efficient verification of computations whose outputs are robust to small changes in oracle answers, achieving near-linear verifier time without relying on cryptography. The authors implement the protocol, demonstrate a preliminary MNIST experiment, and show that an RIP can outperform a state‑of‑the‑art deployed proof system.

benchmarks
More like this →