Efficiently Deploying LLMs with Controlled Risk

• URL: http://arxiv.org/abs/2410.02173v1
• Date: Thu, 3 Oct 2024 03:25:56 GMT
• Title: Efficiently Deploying LLMs with Controlled Risk
• Authors: Michael J. Zellinger, Matt Thomson,
• Abstract summary: We present hierarchical chains with multi-level abstention (HCMA), which use model-intrinsic uncertainty to delegate queries. Our framework presents novel trade-offs between efficiency and risk.
• Score: 0.9208007322096532
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Deploying large language models in production requires simultaneous attention to efficiency and risk control. Prior work has shown the possibility to cut costs while maintaining similar accuracy, but has neglected to focus on risk control. By contrast, here we present hierarchical chains with multi-level abstention (HCMA), which use model-intrinsic uncertainty to delegate queries along the LLM intelligence hierarchy, enabling training-free model switching based solely on black-box API calls. Our framework presents novel trade-offs between efficiency and risk. For example, deploying HCMA on MMLU cuts the error rate of Llama3 405B by 30% when the model is allowed to abstain on 20% of the queries. To calibrate HCMA for optimal performance, our approach uses data-efficient logistic regressions (based on a simple nonlinear feature transformation), which require only 50 or 100 labeled examples to achieve excellent calibration error (ECE), cutting ECE by 50% compared to naive Platt scaling. On free-form generation tasks, we find that chain-of-thought is ineffectual for selective prediction, whereas zero-shot prompting drives error to 0% on TruthfulQA at high abstention rates. As LLMs are increasingly deployed across computing environments with different capabilities (such as mobile, laptop, and cloud), our framework paves the way towards maintaining deployment efficiency while putting in place sharp risk controls.

Related papers

• AegisLLM: Scaling Agentic Systems for Self-Reflective Defense in LLM Security [74.22452069013289]
  AegisLLM is a cooperative multi-agent defense against adversarial attacks and information leakage. We show that scaling agentic reasoning system at test-time substantially enhances robustness without compromising model utility. Comprehensive evaluations across key threat scenarios, including unlearning and jailbreaking, demonstrate the effectiveness of AegisLLM.
  arXiv  Detail & Related papers  (2025-04-29T17:36:05Z)
• SHA256 at SemEval-2025 Task 4: Selective Amnesia -- Constrained Unlearning for Large Language Models via Knowledge Isolation [12.838593066237452]
  Large language models (LLMs) memorize frequently sensitive information during training, posing risks when deploying publicly accessible models. This paper presents our solution to SemEval-2025 Task 4 on targeted unlearning, which combines causal mediation analysis with layer-specific optimization.
  arXiv  Detail & Related papers  (2025-04-17T15:05:40Z)
• DARS: Dynamic Action Re-Sampling to Enhance Coding Agent Performance by Adaptive Tree Traversal [55.13854171147104]
  Large Language Models (LLMs) have revolutionized various domains, including natural language processing, data analysis, and software development. We present Dynamic Action Re-Sampling (DARS), a novel inference time compute scaling approach for coding agents. We evaluate our approach on SWE-Bench Lite benchmark, demonstrating that this scaling strategy achieves a pass@k score of 55% with Claude 3.5 Sonnet V2.
  arXiv  Detail & Related papers  (2025-03-18T14:02:59Z)
• DILEMMA: Joint LLM Quantization and Distributed LLM Inference Over Edge Computing Systems [1.14179290793997]
  This paper introduces DILEMMA, a novel framework addressing the challenges of deploying Large Language Models in Edge Computing systems. DILEMMA formulates an Linear Programming problem to minimize total delay while ensuring acceptable LLM performance levels. It achieves a quantization ratio of up to 12.75% while preserving model loss, highlighting its effectiveness in resource-constrained environments.
  arXiv  Detail & Related papers  (2025-03-03T16:16:33Z)
• Confident or Seek Stronger: Exploring Uncertainty-Based On-device LLM Routing From Benchmarking to Generalization [61.02719787737867]
  Large language models (LLMs) are increasingly deployed and democratized on edge devices. One promising solution is uncertainty-based SLM routing, offloading high-stakes queries to stronger LLMs when resulting in low-confidence responses on SLM. We conduct a comprehensive investigation into benchmarking and generalization of uncertainty-driven routing strategies from SLMs to LLMs over 1500+ settings.
  arXiv  Detail & Related papers  (2025-02-06T18:59:11Z)
• Reward-Guided Speculative Decoding for Efficient LLM Reasoning [80.55186052123196]
  We introduce Reward-Guided Speculative Decoding (RSD), a novel framework aimed at improving the efficiency of inference in large language models (LLMs) RSD incorporates a controlled bias to prioritize high-reward outputs, in contrast to existing speculative decoding methods that enforce strict unbiasedness. RSD delivers significant efficiency gains against decoding with the target model only, while achieving significant better accuracy than parallel decoding method on average.
  arXiv  Detail & Related papers  (2025-01-31T17:19:57Z)
• Activation Sparsity Opportunities for Compressing General Large Language Models [4.5624217435826]
  This work systematically investigates the tradeoff between enforcing activation sparsity and perplexity (accuracy) on state-of-the-art AI models. Our empirical analysis demonstrates that we can obtain around 50% of main memory and computing reductions for critical FFN components with negligible accuracy degradation.
  arXiv  Detail & Related papers  (2024-12-13T02:26:54Z)
• CE-CoLLM: Efficient and Adaptive Large Language Models Through Cloud-Edge Collaboration [1.6021932740447968]
  Large Language Models (LLMs) have achieved remarkable success in serving end-users with human-like intelligence. LLMs demand high computational resources, making it challenging to deploy them to satisfy various performance objectives. We introduce CE-CoLLM, a novel cloud-edge collaboration framework that supports efficient and adaptive LLM inference for end-users at the edge.
  arXiv  Detail & Related papers  (2024-11-05T06:00:27Z)
• Read-ME: Refactorizing LLMs as Router-Decoupled Mixture of Experts with System Co-Design [59.00758127310582]
  We propose a novel framework Read-ME that transforms pre-trained dense LLMs into smaller MoE models. Our approach employs activation sparsity to extract experts. Read-ME outperforms other popular open-source dense models of similar scales.
  arXiv  Detail & Related papers  (2024-10-24T19:48:51Z)
• Attribute Controlled Fine-tuning for Large Language Models: A Case Study on Detoxification [76.14641982122696]
  We propose a constraint learning schema for fine-tuning Large Language Models (LLMs) with attribute control. We show that our approach leads to an LLM that produces fewer inappropriate responses while achieving competitive performance on benchmarks and a toxicity detection task.
  arXiv  Detail & Related papers  (2024-10-07T23:38:58Z)
• Speculative Coreset Selection for Task-Specific Fine-tuning [35.15159197063161]
  Task-specific fine-tuning is essential for the deployment of large language models (LLMs) In this paper, we introduce STAFF, a speculative coreset selection method. We show that STAFF improves the performance of SOTA methods by up to 54.3% and reduces selection overhead by up to 70.5% at different pruning rates.
  arXiv  Detail & Related papers  (2024-10-02T07:42:25Z)
• LoRA-Ensemble: Efficient Uncertainty Modelling for Self-attention Networks [52.46420522934253]
  We introduce LoRA-Ensemble, a parameter-efficient deep ensemble method for self-attention networks. By employing a single pre-trained self-attention network with weights shared across all members, we train member-specific low-rank matrices for the attention projections. Our method exhibits superior calibration compared to explicit ensembles and achieves similar or better accuracy across various prediction tasks and datasets.
  arXiv  Detail & Related papers  (2024-05-23T11:10:32Z)
• FFN-SkipLLM: A Hidden Gem for Autoregressive Decoding with Adaptive Feed Forward Skipping [49.66872823080736]
  Autoregressive Large Language Models (e.g., LLaMa, GPTs) are omnipresent achieving remarkable success in language understanding and generation. To mitigate overload incurred during generation, several early-exit and layer-dropping strategies have been proposed. We propose FFN-SkipLLM, which is an input-adaptive feed-forward skipping strategy.
  arXiv  Detail & Related papers  (2024-04-05T02:35:43Z)
• MobiLlama: Towards Accurate and Lightweight Fully Transparent GPT [87.4910758026772]
  "Bigger the better" has been the predominant trend in recent Large Language Models (LLMs) development. This paper explores the "less is more" paradigm by addressing the challenge of designing accurate yet efficient Small Language Models (SLMs) for resource constrained devices.
  arXiv  Detail & Related papers  (2024-02-26T18:59:03Z)
• Risk-Controlling Model Selection via Guided Bayesian Optimization [35.53469358591976]
  We find a configuration that adheres to user-specified limits on certain risks while being useful with respect to other conflicting metrics. Our method identifies a set of optimal configurations residing in a designated region of interest. We demonstrate the effectiveness of our approach on a range of tasks with multiple desiderata, including low error rates, equitable predictions, handling spurious correlations, managing rate and distortion in generative models, and reducing computational costs.
  arXiv  Detail & Related papers  (2023-12-04T07:29:44Z)
• Cauchy-Schwarz Regularized Autoencoder [68.80569889599434]
  Variational autoencoders (VAE) are a powerful and widely-used class of generative models. We introduce a new constrained objective based on the Cauchy-Schwarz divergence, which can be computed analytically for GMMs. Our objective improves upon variational auto-encoding models in density estimation, unsupervised clustering, semi-supervised learning, and face analysis.
  arXiv  Detail & Related papers  (2021-01-06T17:36:26Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.