MindtheMachine

Research

This page lists formal academic and technical research conducted by Arijit Chatterjee alongside — and independently from — his public writing.

The papers listed here focus on inference-time behavior, internal regulation, and reliability in deployed AI systems.

---

Papers Under Peer Review

(Preprints publicly available via DOI)

Inference-Time Commitment Shaping: A Framework for Quiet Failure Mitigation in LLM Systems

DOI: https://doi.org/10.5281/zenodo.18401073

A technical paper proposing a conceptual and architectural framework for regulating commitment strength in large language model outputs.

---

Control Probe: Inference-Time Commitment Control

DOI: https://doi.org/10.5281/zenodo.18352962

A technical paper examining inference-time commitment control mechanisms that regulate model commitment under evaluative degradation and uncertainty in deployed language models.

---

Evaluative Coherence Regulation (ECR):

An Inference-Time Stability Layer for Enterprise LLMs

DOI: https://doi.org/10.5281/zenodo.18353476

A framework for maintaining evaluative coherence and stability in enterprise-scale large language model deployments through inference-time regulation.

---

Published Working Paper

Need for More Academic R&D in India

DOI: https://doi.org/10.2139/ssrn.1097502

A policy-oriented working paper examining structural gaps and constraints in academic research and development capacity.

---

Scope and Separation

• The papers listed here are technical and academic in nature
• They are independent of public-facing articles and essays
• No Medium articles summarize, derive from, or extend these papers

This separation is intentional.

---

Author

Arijit Chatterjee
ORCID: https://orcid.org/0009-0006-5658-4449