Christopher D. Manning

ICLR Conference 2025 Conference Paper

AuroraCap: Efficient, Performant Video Detailed Captioning and a New Benchmark

• Wenhao Chai
• Enxin Song
• Yilun Du
• Chenlin Meng
• Vashisht Madhavan
• Omer Bar-Tal
• Jenq-Neng Hwang
• Saining Xie

Video detailed captioning is a key task which aims to generate comprehensive and coherent textual descriptions of video content, benefiting both video understanding and generation. In this paper, we propose AuroraCap, a video captioner based on a large multimodal model. We follow the simplest architecture design without additional parameters for temporal modeling. To address the overhead caused by lengthy video sequences, we implement the token merging strategy, reducing the number of input visual tokens. Surprisingly, we found that this strategy results in little performance loss. AuroraCap shows superior performance on various video and image captioning benchmarks, for example, obtaining a CIDEr of 88.9 on Flickr30k, beating GPT-4V (55.3) and Gemini-1.5 Pro (82.2). However, existing video caption benchmarks only include simple descriptions, consisting of a few dozen words, which limits research in this field. Therefore, we develop VDC, a video detailed captioning benchmark with over one thousand carefully annotated structured captions. In addition, we propose a new LLM-assisted metric VDCscore for bettering evaluation, which adopts a divide-and-conquer strategy to transform long caption evaluation into multiple short question-answer pairs. With the help of human Elo ranking, our experiments show that this benchmark better correlates with human judgments of video detailed captioning quality.

ICML Conference 2025 Conference Paper

AxBench: Steering LLMs? Even Simple Baselines Outperform Sparse Autoencoders

• Zhengxuan Wu
• Aryaman Arora
• Atticus Geiger
• Zheng Wang 0078
• Jing Huang 0014
• Dan Jurafsky
• Christopher D. Manning
• Christopher Potts

Fine-grained steering of language model outputs is essential for safety and reliability. Prompting and finetuning are widely used to achieve these goals, but interpretability researchers have proposed a variety of representation-based techniques as well, including sparse autoencoders (SAEs), linear artificial tomography, supervised steering vectors, linear probes, and representation finetuning. At present, there is no benchmark for making direct comparisons between these proposals. Therefore, we introduce AxBench, a large-scale benchmark for steering and concept detection, and report experiments on Gemma-2-2B and 9B. For steering, we find that prompting outperforms all existing methods, followed by finetuning. For concept detection, representation-based methods such as difference-in-means, perform the best. On both evaluations, SAEs are not competitive. We introduce a novel weakly-supervised representational method (Rank-1 Representation Finetuning; ReFT-r1), which is competitive on both tasks while providing the interpretability advantages that prompting lacks. Along with AxBench, we train and publicly release SAE-scale feature dictionaries for ReFT-r1 and DiffMean.

ICLR Conference 2025 Conference Paper

h4rm3l: A Language for Composable Jailbreak Attack Synthesis

• Moussa Koulako Bala Doumbouya
• Ananjan Nandi
• Gabriel Poesia
• Davide Ghilardi
• Anna Goldie
• Federico Bianchi 0001
• Dan Jurafsky
• Christopher D. Manning

Despite their demonstrated valuable capabilities, state-of-the-art (SOTA) widely deployed large language models (LLMs) still have the potential to cause harm to society due to the ineffectiveness of their safety filters, which can be bypassed by prompt transformations called jailbreak attacks. Current approaches to LLM safety assessment, which employ datasets of templated prompts and benchmarking pipelines, fail to cover sufficiently large and diverse sets of jailbreak attacks, leading to the widespread deployment of unsafe LLMs. Recent research showed that novel jailbreak attacks could be derived by composition; however, a formal composable representation for jailbreak attacks, which, among other benefits, could enable the exploration of a large compositional space of jailbreak attacks through program synthesis methods, has not been previously proposed. We introduce h4rm3l, a novel approach that addresses this gap with a human-readable domain-specific language (DSL). Our framework comprises: (1) The h4rm3l DSL, which formally expresses jailbreak attacks as compositions of parameterized string transformation primitives. (2) A synthesizer with bandit algorithms that efficiently generates jailbreak attacks optimized for a target black box LLM. (3) The h4rm3l red-teaming software toolkit that employs the previous two components and an automated harmful LLM behavior classifier that is strongly aligned with human judgment. We demonstrate h4rm3l's efficacy by synthesizing a dataset of 2656 successful novel jailbreak attacks targeting 6 SOTA open-source and proprietary LLMs (GPT-3.5, GPT-4o, Claude-3-Sonnet, Claude-3-Haiku, Llama-3-8B, and Llama-3-70B), and by benchmarking those models against a subset of these synthesized attacks. Our results show that h4rm3l's synthesized attacks are diverse and more successful than existing jailbreak attacks in literature, with success rates exceeding 90% on SOTA LLMs. Warning: This paper and related research artifacts contain offensive and potentially disturbing prompts and model-generated content.

ICLR Conference 2025 Conference Paper

MrT5: Dynamic Token Merging for Efficient Byte-level Language Models

• Julie Kallini
• Shikhar Murty
• Christopher D. Manning
• Christopher Potts
• Róbert Csordás

Models that rely on subword tokenization have significant drawbacks, such as sensitivity to character-level noise like spelling errors and inconsistent compression rates across different languages and scripts. While character- or byte-level models like ByT5 attempt to address these concerns, they have not gained widespread adoption—processing raw byte streams without tokenization results in significantly longer sequence lengths, making training and inference inefficient. This work introduces MrT5 (MergeT5), a more efficient variant of ByT5 that integrates a token deletion mechanism in its encoder to dynamically shorten the input sequence length. After processing through a fixed number of encoder layers, a learned delete gate determines which tokens are to be removed and which are to be retained for subsequent layers. MrT5 effectively "merges" critical information from deleted tokens into a more compact sequence, leveraging contextual information from the remaining tokens. In continued pre-training experiments, we find that MrT5 can achieve significant gains in inference runtime with minimal effect on performance, as measured by bits-per-byte. Additionally, with multilingual training, MrT5 adapts to the orthographic characteristics of each language, learning language-specific compression rates. Furthermore, MrT5 shows comparable accuracy to ByT5 on downstream evaluations such as XNLI, TyDi QA, and character-level tasks while reducing sequence lengths by up to 75%. Our approach presents a solution to the practical limitations of existing byte-level models.

ICLR Conference 2024 Conference Paper

An Emulator for Fine-tuning Large Language Models using Small Language Models

• Eric Mitchell
• Rafael Rafailov
• Archit Sharma
• Chelsea Finn
• Christopher D. Manning

Widely used language models (LMs) are typically built by scaling up a two-stage training pipeline: a pre-training stage that uses a very large, diverse dataset of text and a fine-tuning (sometimes, 'alignment') stage that uses targeted examples or other specifications of desired behaviors. While it has been hypothesized that knowledge and skills come from pre-training, and fine-tuning mostly filters this knowledge and skillset, this intuition has not been extensively tested. To aid in doing so, we introduce a novel technique for decoupling the knowledge and skills gained in these two stages, enabling a direct answer to the question, *What would happen if we combined the knowledge learned by a large model during pre-training with the knowledge learned by a small model during fine-tuning (or vice versa)?* Using an RL-based framework derived from recent developments in learning from human preferences, we introduce *emulated fine-tuning (EFT)*, a principled and practical method for sampling from a distribution that approximates (or 'emulates') the result of pre-training and fine-tuning at different scales. Our experiments with EFT show that scaling up fine-tuning tends to improve helpfulness, while scaling up pre-training tends to improve factuality. Beyond decoupling scale, we show that EFT enables test-time adjustment of competing behavioral traits like helpfulness and harmlessness without additional training. Finally, a special case of emulated fine-tuning, which we call LM *up-scaling*, avoids resource-intensive fine-tuning of large pre-trained models by ensembling them with small fine-tuned models, essentially emulating the result of fine-tuning the large pre-trained model. Up-scaling consistently improves helpfulness and factuality of instruction-following models in the Llama, Llama-2, and Falcon families, without additional hyperparameters or training. For reference implementation, see [https://github.com/eric-mitchell/emulated-fine-tuning](https://github.com/eric-mitchell/emulated-fine-tuning).

ICML Conference 2024 Conference Paper

BAGEL: Bootstrapping Agents by Guiding Exploration with Language

• Shikhar Murty
• Christopher D. Manning
• Peter Shaw 0004
• Mandar Joshi
• Kenton Lee

Following natural language instructions by executing actions in digital environments (e. g. web-browsers and REST APIs) is a challenging task for language model (LM) agents. Unfortunately, LM agents often fail to generalize to new environments without human demonstrations. This work presents BAGEL, a method for bootstrapping LM agents without human supervision. BAGEL converts a seed set of randomly explored trajectories to synthetic demonstrations via round-trips between two noisy LM components: an LM labeler which converts a trajectory into a synthetic instruction, and a zero-shot LM agent which maps the synthetic instruction into a refined trajectory. By performing these round-trips iteratively, BAGEL quickly converts the initial distribution of trajectories towards those that are well-described by natural language. We adapt the base LM agent at test time with in-context learning by retrieving relevant BAGEL demonstrations based on the instruction, and find improvements of over 2-13% absolute on ToolQA and MiniWob++, with up to 13x reduction in execution failures.

ICLR Conference 2024 Conference Paper

Fine-Tuning Language Models for Factuality

• Katherine Tian
• Eric Mitchell
• Huaxiu Yao
• Christopher D. Manning
• Chelsea Finn

The fluency and creativity of large pre-trained language models (LLMs) have led to their widespread use, sometimes even as a replacement for traditional search engines. Yet language models are prone to making convincing but factually inaccurate claims, often referred to as `hallucinations.' These errors can inadvertently spread misinformation or harmfully perpetuate misconceptions. Further, manual fact-checking of model responses is a time-consuming process, making human factuality labels expensive to acquire. In this work, we fine-tune language models to be more factual, without human labeling and targeting more open-ended generation settings than past work. We leverage two key recent innovations in NLP to do so. First, several recent works have proposed methods for judging the factuality of open-ended text by measuring consistency with an external knowledge base or simply a large model's confidence scores. Second, the Direct Preference Optimization algorithm enables straightforward fine-tuning of language models on objectives other than supervised imitation, using a preference ranking over possible model responses. We show that learning from automatically generated factuality preference rankings, generated either through existing retrieval systems or our novel retrieval-free approach, significantly improves the factuality (percent of generated claims that are correct) of Llama-2 on held-out topics compared with RLHF or decoding strategies targeted at factuality. At 7B scale, compared to Llama-2-Chat, we observe 53% and 50% reduction in factual error rate when generating biographies and answering medical questions, respectively. A reference implementation can be found at https://github.com/kttian/llm_factuality_tuning.

ICLR Conference 2024 Conference Paper

Language Model Detectors Are Easily Optimized Against

• Charlotte Nicks
• Eric Mitchell
• Rafael Rafailov
• Archit Sharma
• Christopher D. Manning
• Chelsea Finn
• Stefano Ermon

The fluency and general applicability of large language models (LLMs) has motivated significant interest in detecting whether a piece of text was written by a language model. While both academic and commercial detectors have been deployed in some settings, particularly education, other research has highlighted the fragility of these systems. In this paper, we demonstrate a data-efficient attack that fine-tunes language models to confuse existing detectors, leveraging recent developments in reinforcement learning of language models. We use the `human-ness' score (often just a log probability) of various open-source and commercial detectors as a reward function for reinforcement learning, subject to a KL-divergence constraint that the resulting model does not differ significantly from the original. For a 7B parameter Llama-2 model, fine-tuning for under a day reduces the AUROC of the OpenAI RoBERTa-Large detector from 0.84 to 0.63, while perplexity on OpenWebText increases from 8.7 to only 9.0; with a larger perplexity budget, we can drive AUROC to 0.30 (worse than random). Similar to traditional adversarial attacks, we find that this increase in 'detector evasion' generalizes to other detectors not used during training. In light of our empirical results, we advise against continued reliance on LLM-generated text detectors. Models, datasets, and selected experiment code will be released at https://github.com/charlottttee/llm-detector-evasion.

NeurIPS Conference 2024 Conference Paper

MoEUT: Mixture-of-Experts Universal Transformers

• Róbert Csordás
• Kazuki Irie
• Jürgen Schmidhuber
• Christopher Potts
• Christopher D. Manning

Previous work on Universal Transformers (UTs) has demonstrated the importance of parameter sharing across layers. By allowing recurrence in depth, UTs have advantages over standard Transformers in learning compositional generalizations, but layer-sharing comes with a practical limitation of parameter-compute ratio: it drastically reduces the parameter count compared to the non-shared model with the same dimensionality. Naively scaling up the layer size to compensate for the loss of parameters makes its computational resource requirements prohibitive. In practice, no previous work has succeeded in proposing a shared-layer Transformer design that is competitive in parameter count-dominated tasks such as language modeling. Here we propose MoEUT (pronounced "moot"), an effective mixture-of-experts (MoE)-based shared-layer Transformer architecture, which combines several recent advances in MoEs for both feedforward and attention layers of standard Transformers together with novel layer-normalization and grouping schemes that are specific and crucial to UTs. The resulting UT model, for the first time, slightly outperforms standard Transformers on language modeling tasks such as BLiMP and PIQA, while using significantly less compute and memory.

ICLR Conference 2024 Conference Paper

RAPTOR: Recursive Abstractive Processing for Tree-Organized Retrieval

• Parth Sarthi
• Salman Abdullah
• Aditi Tuli
• Shubh Khanna
• Anna Goldie
• Christopher D. Manning

Retrieval-augmented language models can better adapt to changes in world state and incorporate long-tail knowledge. However, most existing methods retrieve only short contiguous chunks from a retrieval corpus, limiting holistic understanding of the overall document context. We introduce the novel approach of recursively embedding, clustering, and summarizing chunks of text, constructing a tree with differing levels of summarization from the bottom up. At inference time, our RAPTOR model retrieves from this tree, integrating information across lengthy documents at different levels of abstraction. Controlled experiments show that retrieval with recursive summaries offers significant improvements over traditional retrieval-augmented LMs on several tasks. On question-answering tasks that involve complex, multi-step reasoning, we show state-of-the-art results; for example, by coupling RAPTOR retrieval with the use of GPT-4, we can improve the best performance on the QuALITY benchmark by 20\% in absolute accuracy.

NeurIPS Conference 2024 Conference Paper

ReFT: Representation Finetuning for Language Models

• Zhengxuan Wu
• Aryaman Arora
• Zheng Wang
• Atticus Geiger
• Dan Jurafsky
• Christopher D. Manning
• Christopher Potts

Parameter-efficient finetuning (PEFT) methods seek to adapt large neural models via updates to a small number of weights. However, much prior interpretability work has shown that representations encode rich semantic information, suggesting that editing representations might be a more powerful alternative. We pursue this hypothesis by developing a family of Representation Finetuning (ReFT) methods. ReFT methods operate on a frozen base model and learn task-specific interventions on hidden representations. We define a strong instance of the ReFT family, Low-rank Linear Subspace ReFT (LoReFT), and we identify an ablation of this method that trades some performance for increased efficiency. Both are drop-in replacements for existing PEFTs and learn interventions that are 15x--65x more parameter-efficient than LoRA. We showcase LoReFT on eight commonsense reasoning tasks, four arithmetic reasoning tasks, instruction-tuning, and GLUE. In all these evaluations, our ReFTs deliver the best balance of efficiency and performance, and almost always outperform state-of-the-art PEFTs. Upon publication, we will publicly release our generic ReFT training library.

ICLR Conference 2023 Conference Paper

Characterizing intrinsic compositionality in transformers with Tree Projections

• Shikhar Murty
• Pratyusha Sharma
• Jacob Andreas
• Christopher D. Manning

When trained on language data, do transformers learn some arbitrary computation that utilizes the full capacity of the architecture or do they learn a simpler, tree-like computation, hypothesized to underlie compositional meaning systems like human languages? There is an apparent tension between compositional accounts of human language understanding, which are based on a restricted bottom-up computational process, and the enormous success of neural models like transformers, which can route information arbitrarily between different parts of their input. One possibility is that these models, while extremely flexible in principle, in practice learn to interpret language hierarchically, ultimately building sentence representations close to those predictable by a bottom-up, tree-structured model. To evaluate this possibility, we describe an unsupervised and parameter-free method to \emph{functionally project} the behavior of any transformer into the space of tree-structured networks. Given an input sentence, we produce a binary tree that approximates the transformer's representation-building process and a score that captures how ``tree-like'' the transformer's behavior is on the input. While calculation of this score does not require training any additional models, it provably upper-bounds the fit between a transformer and any tree-structured approximation. Using this method, we show that transformers for three different tasks become more tree-like over the course of training, in some cases unsupervisedly recovering the same trees as supervised parsers. These trees, in turn, are predictive of model behavior, with more tree-like models generalizing better on tests of compositional generalization.

ICML Conference 2023 Conference Paper

DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability Curvature

• Eric Mitchell
• Yoonho Lee 0001
• Alexander Khazatsky
• Christopher D. Manning
• Chelsea Finn

The increasing fluency and widespread usage of large language models (LLMs) highlight the desirability of corresponding tools aiding detection of LLM-generated text. In this paper, we identify a property of the structure of an LLM’s probability function that is useful for such detection. Specifically, we demonstrate that text sampled from an LLM tends to occupy negative curvature regions of the model’s log probability function. Leveraging this observation, we then define a new curvature-based criterion for judging if a passage is generated from a given LLM. This approach, which we call DetectGPT, does not require training a separate classifier, collecting a dataset of real or generated passages, or explicitly watermarking generated text. It uses only log probabilities computed by the model of interest and random perturbations of the passage from another generic pre-trained language model (e. g. , T5). We find DetectGPT is more discriminative than existing zero-shot methods for model sample detection, notably improving detection of fake news articles generated by 20B parameter GPT-NeoX from 0. 81 AUROC for the strongest zero-shot baseline to 0. 95 AUROC for DetectGPT.

ICLR Conference 2022 Conference Paper

Fast Model Editing at Scale

• Eric Mitchell
• Charles Lin
• Antoine Bosselut
• Chelsea Finn
• Christopher D. Manning

While large pre-trained models have enabled impressive results on a variety of downstream tasks, the largest existing models still make errors, and even accurate predictions may become outdated over time. Because detecting all such failures at training time is impossible, enabling both developers and end users of such models to correct inaccurate outputs while leaving the model otherwise intact is desirable. However, the distributed, black-box nature of the representations learned by large neural networks makes producing such targeted edits difficult. If presented with only a single problematic input and new desired output, fine-tuning approaches tend to overfit; other editing algorithms are either computationally infeasible or simply ineffective when applied to very large models. To enable easy post-hoc editing at scale, we propose Model Editor Networks using Gradient Decomposition (MEND), a collection of small auxiliary editing networks that use a single desired input-output pair to make fast, local edits to a pre-trained model's behavior. MEND learns to transform the gradient obtained by standard fine-tuning, using a low-rank decomposition of the gradient to make the parameterization of this transformation tractable. MEND can be trained on a single GPU in less than a day even for 10 billion+ parameter models; once trained MEND enables rapid application of new edits to the pre-trained model. Our experiments with T5, GPT, BERT, and BART models show that MEND is the only approach to model editing that effectively edits the behavior of models with more than 10 billion parameters. Code available at https://sites.google.com/view/mend-editing.

ICLR Conference 2022 Conference Paper

GreaseLM: Graph REASoning Enhanced Language Models

• Xikun Zhang 0001
• Antoine Bosselut
• Michihiro Yasunaga
• Hongyu Ren
• Percy Liang
• Christopher D. Manning
• Jure Leskovec

Answering complex questions about textual narratives requires reasoning over both stated context and the world knowledge that underlies it. However, pretrained language models (LM), the foundation of most modern QA systems, do not robustly represent latent relationships between concepts, which is necessary for reasoning. While knowledge graphs (KG) are often used to augment LMs with structured representations of world knowledge, it remains an open question how to effectively fuse and reason over the KG representations and the language context, which provides situational constraints and nuances. In this work, we propose GreaseLM, a new model that fuses encoded representations from pretrained LMs and graph neural networks over multiple layers of modality interaction operations. Information from both modalities propagates to the other, allowing language context representations to be grounded by structured world knowledge, and allowing linguistic nuances (e.g., negation, hedging) in the context to inform the graph representations of knowledge. Our results on three benchmarks in the commonsense reasoning (i.e., CommonsenseQA, OpenbookQA) and medical question answering (i.e., MedQA-USMLE) domains demonstrate that GreaseLM can more reliably answer questions that require reasoning over both situational constraints and structured knowledge, even outperforming models 8x larger.

ICLR Conference 2022 Conference Paper

Hindsight: Posterior-guided training of retrievers for improved open-ended generation

• Ashwin Paranjape
• Omar Khattab
• Christopher Potts
• Matei Zaharia
• Christopher D. Manning

Many text generation systems benefit from retrieving passages from a textual knowledge corpus (e.g., Wikipedia) and using them to generate the output. For open-ended generation tasks, like generating informative utterances in conversations, many varied passages $z$ are relevant to the context $x$ but few are relevant to the observed next utterance $y$ (label). For such tasks, existing methods (that jointly train the retriever and generator) underperform: during training the top-k context-relevant retrieved passages might not contain the label-relevant passage and the generator may hence not learn a preference to ground its generated output in them. We propose using an additional guide-retriever that also conditions on the observed label $y$ and “in hindsight” retrieves label-relevant passages during training. We maximize the evidence lower bound (ELBo) to jointly train the guide-retriever $Q(z|x,y)$ with the standard retriever $P_\eta(z|x)$ and the generator $P_\theta(y|x,z)$ and find that ELBo has better inductive biases than prior work. For informative conversations from the Wizard of Wikipedia dataset, with our posterior-guided training, the retriever finds passages with higher relevance in the top-10 (23% relative improvement), the generator’s responses are more grounded in the retrieved passage (19% relative improvement) and the end-to-end system produces better overall output (6.4% relative improvement).

ICML Conference 2022 Conference Paper

Memory-Based Model Editing at Scale

• Eric Mitchell
• Charles Lin
• Antoine Bosselut
• Christopher D. Manning
• Chelsea Finn

Even the largest neural networks make errors, and once-correct predictions can become invalid as the world changes. Model editors make local updates to the behavior of base (pre-trained) models to inject updated knowledge or correct undesirable behaviors. Existing model editors have shown promise, but also suffer from insufficient expressiveness: they struggle to accurately model an edit’s intended scope (examples affected by the edit), leading to inaccurate predictions for test inputs loosely related to the edit, and they often fail altogether after many edits. As a higher-capacity alternative, we propose Semi-Parametric Editing with a Retrieval-Augmented Counterfactual Model (SERAC), which stores edits in an explicit memory and learns to reason over them to modulate the base model’s predictions as needed. To enable more rigorous evaluation of model editors, we introduce three challenging language model editing problems based on question answering, fact-checking, and dialogue generation. We find that only SERAC achieves high performance on all three problems, consistently outperforming existing approaches to model editing by a significant margin. Code, data, and additional project information will be made available at https: //sites. google. com/view/serac-editing.

AAAI Conference 2022 Conference Paper

Synthetic Disinformation Attacks on Automated Fact Verification Systems

• Yibing Du
• Antoine Bosselut
• Christopher D. Manning

Automated fact-checking is a needed technology to curtail the spread of online misinformation. One current framework for such solutions proposes to verify claims by retrieving supporting or refuting evidence from related textual sources. However, the realistic use cases for fact-checkers will require verifying claims against evidence sources that could be affected by the same misinformation. Furthermore, the development of modern NLP tools that can produce coherent, fabricated content would allow malicious actors to systematically generate adversarial disinformation for fact-checkers. In this work, we explore the sensitivity of automated factcheckers to synthetic adversarial evidence in two simulated settings: ADVERSARIAL ADDITION, where we fabricate documents and add them to the evidence repository available to the fact-checking system, and ADVERSARIAL MODIFICATION, where existing evidence source documents in the repository are automatically altered. Our study across multiple models on three benchmarks demonstrates that these systems suffer significant performance drops against these attacks. Finally, we discuss the growing threat of modern NLG systems as generators of disinformation in the context of the challenges they pose to automated fact-checkers.

ICLR Conference 2020 Conference Paper

ELECTRA: Pre-training Text Encoders as Discriminators Rather Than Generators

• Kevin Clark
• Minh-Thang Luong
• Quoc V. Le
• Christopher D. Manning

Masked language modeling (MLM) pre-training methods such as BERT corrupt the input by replacing some tokens with [MASK] and then train a model to reconstruct the original tokens. While they produce good results when transferred to downstream NLP tasks, they generally require large amounts of compute to be effective. As an alternative, we propose a more sample-efficient pre-training task called replaced token detection. Instead of masking the input, our approach corrupts it by replacing some tokens with plausible alternatives sampled from a small generator network. Then, instead of training a model that predicts the original identities of the corrupted tokens, we train a discriminative model that predicts whether each token in the corrupted input was replaced by a generator sample or not. Thorough experiments demonstrate this new pre-training task is more efficient than MLM because the task is defined over all input tokens rather than just the small subset that was masked out. As a result, the contextual representations learned by our approach substantially outperform the ones learned by BERT given the same model size, data, and compute. The gains are particularly strong for small models; for example, we train a model on one GPU for 4 days that outperforms GPT (trained using 30x more compute) on the GLUE natural language understanding benchmark. Our approach also works well at scale, where it performs comparably to RoBERTa and XLNet while using less than 1/4 of their compute and outperforms them when using the same amount of compute.

ICML Conference 2013 Conference Paper

Fast dropout training

• Sida Wang 0001
• Christopher D. Manning

Preventing feature co-adaptation by encouraging independent contributions from different features often improves classification and regression performance. Dropout training (Hinton et al. , 2012) does this by randomly dropping out (zeroing) hidden units and input features during training of neural networks. However, repeatedly sampling a random subset of input features makes training much slower. Based on an examination of the implied objective function of dropout training, we show how to do fast dropout training by sampling from or integrating a Gaussian approximation, instead of doing Monte Carlo optimization of this objective. This approximation, justified by the central limit theorem and empirical evidence, gives an order of magnitude speedup and more stability. We show how to do fast dropout training for classification, regression, and multilayer neural networks. Beyond dropout, our technique is extended to integrate out other types of noise and small image transformations.

ICML Conference 2013 Conference Paper

Topic Model Diagnostics: Assessing Domain Relevance via Topical Alignment

• Jason Chuang
• Sonal Gupta
• Christopher D. Manning
• Jeffrey Heer

The use of topic models to analyze domain-specific texts often requires manual validation of the latent topics to ensure they are meaningful. We introduce a framework to support large-scale assessment of topical relevance. We measure the correspondence between a set of latent topics and a set of reference concepts to quantify four types of topical misalignment: junk, fused, missing, and repeated topics. Our analysis compares 10, 000 topic model variants to 200 expert-provided domain concepts, and demonstrates how our framework can inform choices of model parameters, inference algorithms, and intrinsic measures of topical quality.

ICML Conference 2011 Conference Paper

Parsing Natural Scenes and Natural Language with Recursive Neural Networks

• Richard Socher
• Cliff Chiung-Yu Lin
• Andrew Y. Ng
• Christopher D. Manning

ICML Conference 2004 Conference Paper

Learning random walk models for inducing word dependency distributions

• Kristina Toutanova
• Christopher D. Manning
• Andrew Y. Ng

IJCAI Conference 2003 Conference Paper

Factored A* Search for Models over Sequences and Trees

• Dan Klein
• Christopher D. Manning

We investigate the calculation of A* bounds for sequence and tree models which are the explicit intersection of a set of simpler models or can be bounded by such an intersection. We provide a natural viewpoint which unifies various instances of factored A* models for trees and sequences, some previously known and others novel, including multiple sequence alignment, weighted finitestate transducer composition, and lexicalized statistical parsing. The specific case of parsing with a product of syntactic (PCFG) and semantic (lexical dependency) components is then considered in detail. We show that this factorization gives a modular lexicalized parser which is simpler than comparably accurate non-factored models, and which allows efficient exact inference with large treebank grammars.

IJCAI Conference 2003 Conference Paper

Spectral Learning

• Sepandar D. Kamvar
• Dan Klein
• Christopher D. Manning

ICML Conference 2002 Conference Paper

From Instance-level Constraints to Space-Level Constraints: Making the Most of Prior Knowledge in Data Clustering

• Dan Klein 0001
• Sepandar D. Kamvar
• Christopher D. Manning

ICML Conference 2002 Conference Paper

Interpreting and Extending Classical Agglomerative Clustering Algorithms using a Model-Based approach

• Sepandar D. Kamvar
• Dan Klein 0001
• Christopher D. Manning