Papers by Thomas Hofmann

5 papers
On the Effect of (Near) Duplicate Subwords in Language Modelling (2024.findings-acl)

Copied to clipboard

Challenge: Tokenisation is a core part of language models but can lead to less efficient training because it removes character-level information.
Approach: They propose to use a tokenisation method to split a character sequence into subwords which are assigned random indices before being served to the LM.
Outcome: The proposed model can generalise across duplicated subwords, but this incurs extra cost and is less data efficient.
Autoregressive Text Generation Beyond Feedback Loops (D19-1)

Copied to clipboard

Challenge: Autoregressive feedback exposes the evolution of the hidden state trajectory to potential biases from well-known train-test discrepancies.
Approach: They combine a latent state space model with a CRF observation model to investigate the state evolution of a hidden state trajectory.
Outcome: The proposed model performs better on unconditional sentence generation compared to baselines while avoiding some prototypical failure modes.
Local and Global Decoding in Text Generation (2024.findings-emnlp)

Copied to clipboard

Challenge: Text generation relies heavily on decoding algorithms that sample strings from a language model distribution.
Approach: They propose to introduce globally-normalised versions of traditional decoding methods and propose an independent Metropolis-Hastings algorithm to approximate sampling from globally-averaged distributions without explicitly computing them.
Outcome: The proposed method approximates the distributions without explicitly computing them.
Decoding a Neural Retriever’s Latent Space for Query Suggestion (2022.emnlp-main)

Copied to clipboard

Challenge: Neural retrieval models have replaced bag-of-words methods for document retrieval . however, they lack the interpretability of bag-off-word models .
Approach: They train a query decoder that generates a meaningful query from a latent representation of a neural search engine.
Outcome: The proposed model outperforms both query reformulation and PRF information retrieval baselines.
Causal Estimation of Tokenisation Bias (2025.acl-long)

Copied to clipboard

Challenge: Modern language models define probabilities over character-strings, but in practice, it does . Ideally, the choice of the tokeniser should not affect the probability assigned to the underlying character- string.
Approach: They quantify a type of tokenisation bias by framing it as a causal effect and estimating it using the regression discontinuity design.
Outcome: The proposed model can estimate tokenisation bias by comparing subwords around arbitrary cutoff points.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations