Instructive Decoding Enhances Instruction-Tuned Language Models without Parameter Updates
Researchers from KAIST AI introduced Instructive Decoding (ID), a method that enhances instruction-tuned LMs without parameter updates.
Instruction-tuned language models (LMs) generalize well to unseen tasks in a zero-shot setting. However, their performance on tasks outside their training data is often limited. Despite being built on large datasets and having billions of parameters, these LMs excel at In-Context Learning (ICL), where they can generate responses to a few examples without needing to be re-trained. However, the training dataset’s scope limits their effectiveness on unfamiliar tasks. Techniques like prompt engineering and output diversification can help improve performance but require significant effort. Recent research explores applying the cognitive anchoring effect to LMs, suggesting that emphasizing initial prompts can enhance task-specific responses and improve fidelity to instructions.
In this work, researchers from KAIST AI introduce Instructive Decoding (ID), a method that enhances instruction-tuned LMs without any parameter updates. Inspired by noisy supervision techniques, ID uses “noisy instructions,” which are altered versions of the original instructions, to create a contrastive approach for predicting the next token. By steering the model’s output in different directions, especially using “opposite” instructions, ID improves model performance across tasks. Experiments show significant gains in accuracy, with smaller models enhanced by ID outperforming larger ones. This method improves adherence to instructions and enhances overall response quality, demonstrating its effectiveness across various models and tasks.
The goal of instruction-tuning is to fine-tune pre-trained LMs to better follow natural language instructions, which improves generalization to unseen tasks, especially in zero-shot scenarios. Expanding the variety and complexity of training tasks enhances this capability, although the models often rely heavily on pre-trained knowledge. Prior research highlights that LMs are sensitive to familiar instructions, even handling misleading ones, and this sensitivity can be leveraged through contrastive techniques. Contrast in text generation, like Contrastive Decoding, compares outputs from different models or inputs to improve performance. This study extends these ideas by using noisy instructions to boost generalization in instruction-tuned LMs.
Instructive Decoding improves response generation in instruction-tuned models by contrasting outputs generated from noisy instructions. It builds on the anchoring effect, where initial information influences subsequent judgments and leverages differences between responses generated from original and altered instructions. The method uses noisy instruction variants like truncated, shuffled, or random words to mislead the model while ensuring task fidelity. By comparing logits from original and noisy instructions during decoding, Instructive Decoding helps models correct biases and produce responses more aligned with the intended instructions, refining their performance on unseen tasks.
The experimental setup uses the SUPNATINST and UNNATINST datasets, evaluating models like Tk-Instruct, Alpaca, and T0 across tasks like Grammar Error Correction and Textual Entailment. Rouge-L, Exact Match (EM), Label Adherence (LA), and Label Coherence (LC) metrics assess performance. ID consistently improves results, especially for larger models like Tk-XXL, enhancing LA and LC. Interestingly, noisy instructions enhance output quality with ID despite baseline performance degradation. Though task-specific performance varies, the ‘opposite’ instruction variant proves robust across tasks. Overall, ID shows significant gains across model sizes and task types.
The study investigates the challenges of unseen task generalization in instruction-tuned language models. The proposed method, ID, leverages the anchoring effect using “noisy” instructions to counteract inherent model biases. By contrasting predictions with those generated from altered instructions, ID enhances model performance, particularly with the “opposite” noisy variant, which deviates most from the original input. Empirical results show ID’s effectiveness across multiple tasks, with notable improvements in prediction diversity. The approach requires no additional parameter updates, making it a practical tool for improving instruction-following in language models.
Check out the Paper. All credit for this research goes to the researchers of this project. Also, don’t forget to follow us on Twitter and join our Telegram Channel and LinkedIn Group. If you like our work, you will love our newsletter.
Don’t Forget to join our 50k ML SubReddit
The above is the detailed content of Instructive Decoding Enhances Instruction-Tuned Language Models without Parameter Updates. For more information, please follow other related articles on the PHP Chinese website!

FloppyPepe (FPPE) is set to soar from $0.0000002 to $1, with experts claiming it could cancel six zeroes before Shiba Inu (SHIB) and Pepe (PEPE) make their next move.

Every crypto cycle brings a fresh batch of outrageous, community-fueled tokens that take the market by storm, flipping wallets overnight and minting new millionaires.

When Bonk launched as the Solana ecosystem's first meme coin, few saw it coming. Traded for fractions of a cent in its infancy, Bonk was written off by many as just another

Binance Wallet has rolled out support for Sonic Chain, a rebranded version of the Fantom blockchain, marking a new chapter for the network led by DeFi pioneer Andre Cronje.

The legality of sports betting in Australia is a complex topic. Online sports betting is legal and heavily regulated

Since the recent market correction, Bitcoin's long-term holders have increased their supply by approximately 254,000 BTC. This trend reflects rising confidence as large holders accumulate at higher price levels.

After surging into the $97,000 level this week, Bitcoin appears to be entering a fresh bullish phase.

On April 29, 2025, the company entered into a deal with an unidentified institutional partner to secure financing of up to $20 million.

Hot AI Tools

Undresser.AI Undress
AI-powered app for creating realistic nude photos

AI Clothes Remover
Online AI tool for removing clothes from photos.

Undress AI Tool
Undress images for free

Clothoff.io
AI clothes remover

Video Face Swap
Swap faces in any video effortlessly with our completely free AI face swap tool!

Hot Article

Hot Tools

SublimeText3 Chinese version
Chinese version, very easy to use

MinGW - Minimalist GNU for Windows
This project is in the process of being migrated to osdn.net/projects/mingw, you can continue to follow us there. MinGW: A native Windows port of the GNU Compiler Collection (GCC), freely distributable import libraries and header files for building native Windows applications; includes extensions to the MSVC runtime to support C99 functionality. All MinGW software can run on 64-bit Windows platforms.

Safe Exam Browser
Safe Exam Browser is a secure browser environment for taking online exams securely. This software turns any computer into a secure workstation. It controls access to any utility and prevents students from using unauthorized resources.

SecLists
SecLists is the ultimate security tester's companion. It is a collection of various types of lists that are frequently used during security assessments, all in one place. SecLists helps make security testing more efficient and productive by conveniently providing all the lists a security tester might need. List types include usernames, passwords, URLs, fuzzing payloads, sensitive data patterns, web shells, and more. The tester can simply pull this repository onto a new test machine and he will have access to every type of list he needs.

SAP NetWeaver Server Adapter for Eclipse
Integrate Eclipse with SAP NetWeaver application server.
