English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

A large body of NLP research has documented the ways gender biases manifestand amplify within large language models (LLMs), though this research haspredominantly operated within a gender binary-centric context. A growing bodyof work has identified the harmful limitations of this gender-exclusiveframing; many LLMs cannot correctly and consistently refer to persons outsidethe gender binary, especially if they use neopronouns. While data scarcity hasbeen identified as a possible culprit, the precise mechanisms through which itinfluences LLM misgendering remain underexplored. Our work addresses this gapby studying data scarcity's role in subword tokenization and, consequently, theformation of LLM word representations. We uncover how the Byte-Pair Encoding(BPE) tokenizer, a backbone for many popular LLMs, contributes to neopronounmisgendering through out-of-vocabulary behavior. We introduce pronountokenization parity (PTP), a novel approach to reduce LLM neopronounmisgendering by preserving a token's functional structure. We evaluate PTP'sefficacy using pronoun consistency-based metrics and a novel syntax-basedmetric. Through several controlled experiments, finetuning LLMs with PTPimproves neopronoun consistency from 14.5% to 58.4%, highlighting thesignificant role tokenization plays in LLM pronoun consistency.

Are you talking to ['xem'] or ['x', 'em']? On Tokenization and Addressing Misgendering in LLMs with Pronoun Tokenization Parity