Aging with GRACE: Lifelong Model Editing with Discrete Key-Value Adaptors
Reviewed by Pith T0 review T1 audit T2 compute T3 formal T4 kernel pith:Z2BXPRTYrecord.jsonopen to challenge →
read the original abstract
Deployed language models decay over time due to shifting inputs, changing user needs, or emergent world-knowledge gaps. When such problems are identified, we want to make targeted edits while avoiding expensive retraining. However, current model editors, which modify such behaviors of pre-trained models, degrade model performance quickly across multiple, sequential edits. We propose GRACE, a lifelong model editing method, which implements spot-fixes on streaming errors of a deployed model, ensuring minimal impact on unrelated inputs. GRACE writes new mappings into a pre-trained model's latent space, creating a discrete, local codebook of edits without altering model weights. This is the first method enabling thousands of sequential edits using only streaming errors. Our experiments on T5, BERT, and GPT models show GRACE's state-of-the-art performance in making and retaining edits, while generalizing to unseen inputs. Our code is available at https://www.github.com/thartvigsen/grace}.
This paper has not been read by Pith yet.
Forward citations
Cited by 12 Pith papers
-
When to Write and When to Suppress: Route-Specialized Dual Adapters for Memory-Assisted Knowledge Editing
RRDA introduces a router plus separate edit and locality adapters for memory-assisted knowledge editing, reporting highest accuracy on CounterFact, ZsRE, and MQuAKE-CF across two 8B models.
-
TimeROME-DLM: Temporal Causal Tracing and Low-Rank Inference-Time Knowledge Editing for Masked Diffusion Language Models
TimeROME-DLM enables training-free knowledge editing in masked diffusion language models via temporal causal tracing and low-rank residual edit memory applied at inference time.
-
Knowledge Editing in Masked Diffusion Language Models
Locate-then-edit succeeds at the same early-to-mid MLP locations in masked diffusion models as in autoregressive models, but requires optimization over intermediate partial-mask states to handle multi-token targets.
-
Beyond Binary Edits Robust Multimodal Knowledge Editing with Adversarial Subspace Alignment
Introduces Latent Adversarial Robustification and Rank-Constrained Subspace Learning to enable robust generalization in multimodal knowledge editing through adversarial subspace alignment.
-
Norm Anchors Make Model Edits Last
Norm-Anchor Scaling breaks the norm-feedback loop in sequential LLM editing by anchoring value vectors to original norms, improving long-run performance by 72.2% and extending the editing horizon over 4x.
-
Evaluating and Understanding Model Editing for Medical Vision Language Models
M3Bench is a clinically grounded benchmark showing that gradient-based VLM editors generalize but break locality, while memory-based editors preserve locality but fail on composition and temporal tasks, with failures ...
-
Memory Depth, Not Memory Access: Selective Parametric Consolidation for Long-Running Language Agents
EVAF, a surprise- and valence-gated LoRA mechanism, provides memory depth for goal persistence in language agents via the loop-drift protocol, complementary to retrieval.
-
When to Write and When to Suppress: Route-Specialized Dual Adapters for Memory-Assisted Knowledge Editing
Introduces route-specialized dual adapters that route prompts to either an edit adapter or a locality adapter, achieving highest accuracy on CF, ZSRE, and MQuAKE benchmarks for 7B/8B models.
-
Diffusion-Inspired Masked Fine-Tuning for Knowledge Injection in Autoregressive LLMs
Masked fine-tuning enables autoregressive LLMs to inject new factual knowledge without paraphrases and with reversal-curse resistance, matching diffusion LLM advantages on QA tasks.
-
EVAF: A Test-Retest Protocol for Selective Parametric Consolidation
EVAF and test-retest protocol show selective parametric consolidation of high-valence experiences in GPT-2 and TinyLlama while preserving factual retrieval.
-
Towards Scalable Lifelong Knowledge Editing with Selective Knowledge Suppression
LightEdit enables scalable lifelong knowledge editing in LLMs via selective knowledge retrieval and probability suppression during decoding, outperforming prior methods on ZSRE, Counterfact, and RIPE while reducing tr...
-
MemOS: A Memory OS for AI System
MemOS introduces a unified memory management framework for LLMs using MemCubes to handle and evolve different memory types for improved controllability and evolvability.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.