14 papers
Prompt-Injection Probes Learn the Conversation, Not Just the Injection
Kaan Kale, Sriram Vishwanath
Linear probes on a model's hidden states are used as cheap monitors for indirect prompt injection: they screen the tool outputs an agent reads and send suspicious ones to a slower…
AIProver: Agentic Auto-Formalization of Mathematical Research via Certificate-Driven Evolving Harness
Prithwish Jana, Viet Bach Hoang, Logan Luna +9
Proof auto-formalization translates natural-language (NL) theorems and proofs into a formal language (FL) such as Lean, enabling mechanical verification. Despite rapid progress, re…
WaveCommitments: Private and Verifiable Speech-Agent Marketplaces
Oguzhan Baser, Kaan Kale, Sriram Vishwanath +1
Modern speech systems are evolving into ecosystems of autonomous speech agents that interact, negotiate, and trade assets such as embeddings, spectrograms, or raw data. In such mar…
From Verification Failures to Reusable Guidance for Coding Agents
Yuqing Zhai, Xiaohong Chen, Lingming Zhang +2
Coding agents need to establish that a program satisfies a specification and that the specification captures the requested behavior. We study how expert diagnosis of verification f…
ModalFidelity: Routing Modalities for Deepfake Detection on a Budget
Oguzhan Baser, Kaan Kale, Sriram Vishwanath +1
Deepfakes no longer need to fake a whole video. Generators that read the transcript now alter only the few seconds in which a video's meaning turns, so a forgery hides in a small,…
The Key Handoff: Retrieval in Hybrid Language Models
Kaan Kale, Oguzhan Baser, Sriram Vishwanath
A two-hop question makes a language model retrieve twice: once to produce a bridge entity, and once to retrieve with it. Transformers resolve that entity in their early layers. Hyb…
GT-VLA: Target-Conditioned Trace Guidance for Generalizable Robotic Manipulation
Ninghan Zhong, Jing-Chen Peng, Sriram Vishwanath
Vision-Language-Action (VLA) models have shown strong performance on robotic manipulation, but they often struggle to generalize to unseen tasks, configurations, and long-horizon s…
The Verbose Context Problem in Medical Records
Shiva Kaul, Min-Gyu Kim, Anjum Khurshid +1
The verbose context problem occurs when structured concepts have token-inefficient textual representations. This bottleneck is acute in population health: cohort-level analysis of…
VeryTrace: Verifying Reasoning Traces through Compilable Formalism and Structured Verification
Ninghan Zhong, Ahmet Ege Tanriverdi, Kaan Kale +1
Multi-step reasoning with Chain-of-Thought (CoT) prompting remains fragile: logical errors or hallucinations in early steps silently propagate, producing confident but incorrect co…
Symbolic Regression via Latent Iterative Refinement
Xieting Chu, Sriram Vishwanath, Vijay Ganesh
Symbolic regression (SR) seeks closed-form mathematical expressions that fit observed data. Neural SR methods amortize the search by training an encoder to map observations directl…
TensorCommitments: A Lightweight Verifiable Inference for Large Language Models
Oguzhan Baser, Elahe Sadeghi, Eric Wang +6
Most large language models (LLMs) run on external clouds: users send a prompt, pay for inference, and must trust that the remote GPU executes the LLM without any adversarial tamper…
OPTIMUM-DERAM: Highly Consistent, Scalable, and Secure Multi-Object Memory using RLNC
Nicolas Nicolaou, Kishori M. Konwar, Moritz Grundei +3
This paper introduces OPTIMUM-DERAM, a highly consistent, scalable, secure, and decentralized shared memory solution. Traditional distributed shared memory implementations offer mu…
ProofBridge: Auto-Formalization of Natural Language Proofs in Lean via Joint Embeddings
Prithwish Jana, Kaan Kale, Ahmet Ege Tanriverdi +3
Translating human-written mathematical theorems and proofs from natural language (NL) into formal languages (FLs) like Lean 4 has long been a significant challenge for AI. Most sta…
OPTIMUMP2P: Fast and Reliable Gossiping in P2P Networks
Nicolas Nicolaou, Onyeka Obi, Aayush Rajasekaran +8
Gossip algorithms are pivotal in the dissemination of information within decentralized systems. Consequently, numerous gossip libraries have been developed and widely utilized espe…