Pith. sign in

REVIEW 10 cited by

InstructRetro: Instruction Tuning post Retrieval-Augmented Pretraining

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2310.07713 v3 pith:TY53PNBR submitted 2023-10-11 cs.CL cs.AIcs.IRcs.LG

Boxin Wang , Wei Ping , Lawrence McAfee , Peng Xu , Bo Li , Mohammad Shoeybi , Bryan Catanzaro This is my paper
classification cs.CLcs.AIcs.IRcs.LG
keywords instructionretroinstructretrotaskstuningacrosspretrainingretrieval
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Pretraining auto-regressive large language models~(LLMs) with retrieval demonstrates better perplexity and factual accuracy by leveraging external databases. However, the size of existing pretrained retrieval-augmented LLM is still limited (e.g., Retro has 7.5B parameters), which limits the effectiveness of instruction tuning and zero-shot generalization. In this work, we introduce Retro 48B, the largest LLM pretrained with retrieval. Specifically, we continue to pretrain a 43B GPT model on additional 100 billion tokens using the Retro augmentation method by retrieving from 1.2 trillion tokens. Notably, the obtained foundation model, Retro 48B, largely outperforms the counterpart GPT 43B trained on 1.2T tokens in terms of perplexity with only 2.58% additional GPU hours, demonstrating the significant scaling potential of the method. After instruction tuning on Retro, InstructRetro demonstrates significant improvement over the instruction tuned GPT on a wide range of zero-shot tasks. Specifically, the average improvement of InstructRetro is 7% over its GPT counterpart across 8 short-form QA and reading comprehension tasks, 10% over GPT across 4 challenging long-form QA tasks, and 16% over GPT across 3 summarization tasks. Surprisingly, we find that one can ablate the encoder from InstructRetro architecture and directly use its decoder backbone, while achieving comparable results. Our results highlight the promising direction to obtain a better GPT decoder through continued pretraining with retrieval before instruction tuning. Our code and checkpoints are publicly available at: https://huggingface.co/nvidia/retro-48b-instruct-4k.

Discussion (0). Sign in to comment.

Forward citations

Cited by 10 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. REALISTA: Realistic Latent Adversarial Attacks that Elicit LLM Hallucinations

    cs.CL 2026-05 unverdicted novelty 8.0 of 10

    REALISTA optimizes continuous combinations of valid editing directions in latent space to produce realistic adversarial prompts that elicit hallucinations more effectively than prior methods, including on large reason...

  2. Retrieval from Within: An Intrinsic Capability of Attention-Based Models

    cs.LG 2026-05 unverdicted novelty 7.0 of 10

    Attention-based models can intrinsically retrieve and reuse pre-encoded evidence chunks via decoder attention queries, unifying retrieval with generation and outperforming external RAG pipelines on QA benchmarks.

  3. MMSearch-R1: Incentivizing LMMs to Search

    cs.CV 2025-06 unverdicted novelty 7.0 of 10

    MMSearch-R1 uses reinforcement learning to train multimodal models for on-demand multi-turn internet search with image and text tools, outperforming same-size RAG baselines and matching larger ones while cutting searc...

  4. REALISTA: Realistic Latent Adversarial Attacks that Elicit LLM Hallucinations

    cs.CL 2026-05 unverdicted novelty 6.0 of 10

    REALISTA generates semantically coherent adversarial prompts via latent-space optimization over input-dependent editing directions, achieving stronger hallucination elicitation than prior realistic attacks on open-sou...

  5. Retrieval from Within: An Intrinsic Capability of Attention-Based Models

    cs.LG 2026-05 unverdicted novelty 6.0 of 10

    Attention-based models can retrieve evidence intrinsically by using decoder attention to score and reuse their own pre-encoded chunks, outperforming separate retrieval pipelines on QA benchmarks.

  6. SearchInstruct: Enhancing Domain Adaptation via Retrieval-Based Instruction Dataset Creation

    cs.CL 2025-09 conditional novelty 6.0 of 10

    SearchInstruct expands seed questions with an LLM, retrieves relevant documents, and grounds answers in them, improving domain-specific SFT and enabling targeted model updates.

  7. HIRAG: Hierarchical-Thought Instruction-Tuning Retrieval-Augmented Generation

    cs.CL 2025-07 conditional novelty 6.0 of 10

    A hierarchical chain-of-thought instruction-tuning curriculum for filtering, combination, and reasoning improves zero-shot retrieval-augmented QA.

  8. Retrieval-Augmented Generation for Natural Language Processing: A Survey

    cs.CL 2024-07 accept novelty 6.0 of 10

    The survey organizes RAG methods via a taxonomy of query-based, logits-based, latent, and parametric fusion with comparisons on accessibility, efficiency, applications, and challenges.

  9. NV-Embed: Improved Techniques for Training LLMs as Generalist Embedding Models

    cs.CL 2024-05 accept novelty 6.0 of 10

    NV-Embed achieves first place on the MTEB leaderboard across 56 tasks by combining a latent attention layer, causal-mask removal, two-stage contrastive training, and data curation for LLM-based embedding models.

  10. Retrieval-Augmented Generation for Large Language Models: A Survey

    cs.CL 2023-12 unverdicted novelty 3.0 of 10

    A survey of RAG paradigms, components, benchmarks, and challenges for improving LLMs on knowledge-intensive tasks.

Pith tools