Let's Think Dot by Dot: Hidden Computation in Transformer Language Models

• URL: http://arxiv.org/abs/2404.15758v1
• Date: Wed, 24 Apr 2024 09:30:00 GMT
• Title: Let's Think Dot by Dot: Hidden Computation in Transformer Language Models
• Authors: Jacob Pfau, William Merrill, Samuel R. Bowman,
• Abstract summary: Chain-of-thought responses from language models improve performance across most benchmarks. We show that transformers can use meaningless filler tokens in place of a chain of thought to solve two hard algorithmic tasks. We find that learning to use filler tokens is difficult and requires specific, dense supervision to converge.
• Score: 30.972412126012884
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Chain-of-thought responses from language models improve performance across most benchmarks. However, it remains unclear to what extent these performance gains can be attributed to human-like task decomposition or simply the greater computation that additional tokens allow. We show that transformers can use meaningless filler tokens (e.g., '......') in place of a chain of thought to solve two hard algorithmic tasks they could not solve when responding without intermediate tokens. However, we find empirically that learning to use filler tokens is difficult and requires specific, dense supervision to converge. We also provide a theoretical characterization of the class of problems where filler tokens are useful in terms of the quantifier depth of a first-order formula. For problems satisfying this characterization, chain-of-thought tokens need not provide information about the intermediate computational steps involved in multi-token computations. In summary, our results show that additional tokens can provide computational benefits independent of token choice. The fact that intermediate tokens can act as filler tokens raises concerns about large language models engaging in unauditable, hidden computations that are increasingly detached from the observed chain-of-thought tokens.

Related papers

• Fast Controlled Generation from Language Models with Adaptive Weighted Rejection Sampling [90.86991492288487]
  evaluating constraint on every token can be prohibitively expensive. LCD can distort the global distribution over strings, sampling tokens based only on local information. We show that our approach is superior to state-of-the-art baselines.
  arXiv  Detail & Related papers  (2025-04-07T18:30:18Z)
• How Well do LLMs Compress Their Own Chain-of-Thought? A Token Complexity Approach [4.055489363682199]
  We conduct the first systematic study of the relationship between reasoning length and model performance. We show that this tradeoff persists across even very distinct reasoning chains. We show that prompt-based compression strategies operate far from theoretical limits.
  arXiv  Detail & Related papers  (2025-03-03T03:48:20Z)
• Token Pruning in Multimodal Large Language Models: Are We Solving the Right Problem? [19.35502303812707]
  Multimodal large language models (MLLMs) have shown remarkable performance for cross-modal understanding and generation, yet still suffer from severe inference costs. Recently, abundant works have been proposed to solve this problem with token pruning, which identifies the redundant tokens in MLLMs and then prunes them to reduce the computation and KV storage costs. In this paper, we answer these questions one by one, providing insights into the design of future token pruning methods.
  arXiv  Detail & Related papers  (2025-02-17T07:05:36Z)
• Stop Looking for Important Tokens in Multimodal Language Models: Duplication Matters More [18.928285521147057]
  We show that importance is not an ideal indicator to decide whether a token should be pruned. We propose DART (Duplication-Aware Reduction of Tokens), which prunes tokens based on its duplication with other tokens. Experiments demonstrate that DART can prune 88.9% vision tokens while maintaining comparable performance.
  arXiv  Detail & Related papers  (2025-02-17T06:56:28Z)
• Disentangling Reasoning Tokens and Boilerplate Tokens For Language Model Fine-tuning [46.43130011147807]
  We argue that tokens serving different roles - specifically, reasoning tokens versus boilerplate tokens - differ significantly in importance and learning complexity. We propose a novel Shuffle-Aware Discriminator (SHAD) for adaptive token discrimination. Using SHAD, we propose the Reasoning-highlighted Fine-Tuning (RFT) method, which adaptively emphasizes reasoning tokens during fine-tuning.
  arXiv  Detail & Related papers  (2024-12-19T12:06:24Z)
• SepLLM: Accelerate Large Language Models by Compressing One Segment into One Separator [65.62084602011596]
  Large Language Models (LLMs) have exhibited exceptional performance across a spectrum of natural language processing tasks. We have identified a key pattern: certain seemingly meaningless separator tokens (i.e., punctuations) contribute disproportionately to attention scores compared to semantically meaningful tokens. We introduce SepLLM, a plug-and-play framework that accelerates inference by compressing these segments and eliminating redundant tokens.
  arXiv  Detail & Related papers  (2024-12-16T18:58:57Z)
• Improbable Bigrams Expose Vulnerabilities of Incomplete Tokens in Byte-Level Tokenizers [32.274579719726546]
  Tokenization is a crucial step that bridges human-readable text with model-readable discrete tokens. Recent studies have revealed that tokenizers can be exploited to elicit unwanted model behaviors. We investigate incomplete tokens, i.e., undecodable tokens with stray bytes resulting from byte-level byte-pair encoding (BPE) tokenization.
  arXiv  Detail & Related papers  (2024-10-31T07:19:44Z)
• ElasticTok: Adaptive Tokenization for Image and Video [109.75935878130582]
  We introduce ElasticTok, a method that conditions on prior frames to adaptively encode a frame into a variable number of tokens. During inference, ElasticTok can dynamically allocate tokens when needed. Our evaluations on images and video demonstrate the effectiveness of our approach in efficient token usage.
  arXiv  Detail & Related papers  (2024-10-10T20:54:15Z)
• ToSA: Token Selective Attention for Efficient Vision Transformers [50.13756218204456]
  ToSA is a token selective attention approach that can identify tokens that need to be attended as well as those that can skip a transformer layer. We show that ToSA can significantly reduce computation costs while maintaining accuracy on the ImageNet classification benchmark.
  arXiv  Detail & Related papers  (2024-06-13T05:17:21Z)
• SEP: Self-Enhanced Prompt Tuning for Visual-Language Model [93.94454894142413]
  We introduce a novel approach named Self-Enhanced Prompt Tuning (SEP) SEP explicitly incorporates discriminative prior knowledge to enhance both textual-level and visual-level embeddings. Comprehensive evaluations across various benchmarks and tasks confirm SEP's efficacy in prompt tuning.
  arXiv  Detail & Related papers  (2024-05-24T13:35:56Z)
• EMS-SD: Efficient Multi-sample Speculative Decoding for Accelerating Large Language Models [40.651650382105636]
  Vanilla method adds padding tokens in order to ensure that the number of new tokens remains consistent across samples. We propose a novel method that can resolve the issue of inconsistent tokens accepted by different samples without necessitating an increase in memory or computing overhead. Our proposed method can handle the situation where the prediction tokens of different samples are inconsistent without the need to add padding tokens.
  arXiv  Detail & Related papers  (2024-05-13T08:24:21Z)
• Tokenization counts: the impact of tokenization on arithmetic in frontier LLMs [3.6722413665749674]
  Tokenization is the division of input text into input tokens. We study the effect this choice has on numerical reasoning through the use of arithmetic tasks.
  arXiv  Detail & Related papers  (2024-02-22T18:14:09Z)
• Identifying and Analyzing Task-Encoding Tokens in Large Language Models [55.03191279766383]
  In this paper, we identify and analyze task-encoding tokens on whose representations the task performance depends. We show that template and stopword tokens are the most prone to be task-encoding. Our work sheds light on how large language models (LLMs) learn to perform a task from demonstrations, deepens our understanding of the varied roles different types of tokens play in LLMs, and provides insights for avoiding instability from improperly utilizing task-encoding tokens.
  arXiv  Detail & Related papers  (2024-01-20T20:55:21Z)
• Dynamic Token Pruning in Plain Vision Transformers for Semantic Segmentation [18.168932826183024]
  This work introduces a Dynamic Token Pruning (DToP) method based on the early exit of tokens for semantic segmentation. Experiments suggest that the proposed DToP architecture reduces on average $20% - 35%$ of computational cost for current semantic segmentation methods.
  arXiv  Detail & Related papers  (2023-08-02T09:40:02Z)
• Improving Tokenisation by Alternative Treatment of Spaces [7.596737214110957]
  We experiment with an alternative tokenisation approach where spaces are always treated as individual tokens. We find that our modified algorithms lead to improved performance on downstream NLP tasks.
  arXiv  Detail & Related papers  (2022-04-08T13:22:30Z)
• Fast End-to-End Speech Recognition via a Non-Autoregressive Model and Cross-Modal Knowledge Transferring from BERT [72.93855288283059]
  We propose a non-autoregressive speech recognition model called LASO (Listen Attentively, and Spell Once) The model consists of an encoder, a decoder, and a position dependent summarizer (PDS)
  arXiv  Detail & Related papers  (2021-02-15T15:18:59Z)
• Non-Autoregressive Machine Translation with Disentangled Context Transformer [70.95181466892795]
  State-of-the-art neural machine translation models generate a translation from left to right and every step is conditioned on the previously generated tokens. We propose an attention-masking based model, called Disentangled Context (DisCo) transformer, that simultaneously generates all tokens given different contexts. Our model achieves competitive, if not better, performance compared to the state of the art in non-autoregressive machine translation while significantly reducing decoding time on average.
  arXiv  Detail & Related papers  (2020-01-15T05:32:18Z)

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.