ICML 2026 Past Large language modelsSafety & alignment
The Second Workshop on the Impact of Memorization on Trustworthy Foundation Models at ICML
ICML MemFM 2026 Workshop
- Submission deadline
- May 9, 2026, 12:00 UTC imported from OpenReview — check the website for extensions
- Submission portal
- OpenReview
- Notes
- Auto-imported from the OpenReview venue record on 2026-06-10 — please verify and enrich (topics are keyword-guessed).
Accepted papers (45)
Fetched from OpenReview (v2) on 2026-06-10.
-
\textsc{ContinuousBench}: Can Differentially Private Synthetic Text Improve Capabilities?
-
Alignment Whack-a-Mole : Finetuning Activates Verbatim Recall of Copyrighted Books in Large Language Models
-
Alignment-aware Data Selection for Unlearning in Contrastive Vision-Language Models
-
Amplifying Membership Signal Through Iterative Regeneration
-
An Explicit Memory-Driven Agentic Framework for Power System Simulation
-
Auditing Reasoning-Trace Memorization Claims after Unlearning with Head-Conditioned Canaries
-
Bayes-Optimal Coexistence via Fact Localizability in Trainable-Feature Decoder-Only Transformers
-
Break the Output Geometry for Large Language Model Unlearning
-
Cheap Forgetting: Linear Adapter Interpolation as a Post-Hoc Memorization Mitigation
-
Deployment-Time Memorization in Foundation-Model Agents
-
Detecting Functional Memorization in Code Language Models
-
Do Text Anonymizers Generalize Across Contexts? Extending RAT-Bench to Malaysian Microdata and PII
-
Estimating Model-Level Membership Inference Vulnerability Without Reference Models
-
Estimating near-verbatim extraction risk in language models with decoding-constrained beam search
-
Evidence-bearing Insights under Differential Privacy: Beyond the Limits of Private Text Generation
-
Internal Data Repetition Destroys Language Models
-
KVEraser: Learning to Steer KV Cache for Efficient Localized Context Erasing
-
Local Coverage Governs Memorization in Diffusion Models
-
Machine Text Detectors are Membership Inference Attacks
-
MemBoost: A Memory-Boosted Framework for Cost-Aware LLM Inference
-
Memorization Dynamics of Fill-in-the-Middle Pretraining
-
Memorization Removal as a Two-Player Game: The Adversarial Work Criterion as a Test for Foundation-Model Defenses
-
Memory Adapters Enable Fast, Flexible Knowledge Unlearning in LLMs
-
Mitigating Unintended Memory Use in LLMs via Structured Memory
-
NumLeak: Public Numeric Benchmarks as Latent Label in Foundation Models
-
On Optimization Complexity of Second-Order Certified Unlearning
-
On the Geometry of Memorization: Interpolation and Second-Order Representation Irregularity
-
On the Learning Dynamics of Label-Noise Memorization in ReLU MLPs
-
Position: The Term “Machine Unlearning” Is Overused in LLMs
-
Probing Memorization of Tabular In-Context Learning
-
Probing Policy-Level Memorization in Reasoning LLMs via Atomic Chess
-
Prune to Protect: Faster Training and Enhanced Privacy by Dynamic Data Pruning
-
Rare, Distinctive, Memorized: Auditing Memorization in Fine-Tuned Medical Foundation Models
-
Reconstructing Training Images from Foundation Model Parameters in the Healthcare Domain: Privacy Risks and Defences
-
Scale Dependent Data Duplication
-
Semantic Gravity: When Parametric Memory Overpowers Visual Thermodynamics in Video-LLMs
-
Structural Memorization in AlphaFold: Adversarial Mutations Reveal Template Reliance, Confidence Failures, and Implications for Protein Design
-
Suppression is not Deletion: Adversarial Probes Recover Unlearned Knowledge in Code LLMs
-
SYMBOLICDRIFT: Measuring Reasoning Drift on Unverifiable Questions
-
Synthetic Data and the Rise of Spiky Intelligence
-
The Distillation Game: Adaptive Attacks & Efficient Defenses
-
The Source of Competence Shapes Metacognition in Language Models
-
Watermarking for Proprietary Dataset Protection
-
What to Forget in Unlearning? Forget Set Curation for Language Models
-
Why Forget-Only Unlearning Needs Memorization