The field of large language models (LLMs) is rapidly advancing, with a growing focus on addressing the risks of misinformation and bias. Recent research has highlighted the potential for LLMs to amplify and reinforce existing biases, as well as generate persuasive but misleading content. To combat these issues, researchers are exploring various approaches, including the development of counterspeech generation models, adversarial training frameworks, and fact-checking systems. Notably, some studies have introduced novel frameworks for detecting and mitigating bias in LLMs, such as the use of Bayesian rationality and symbolic adversarial learning. Furthermore, researchers are investigating the impact of LLMs on social dynamics, including the spread of misinformation and the erosion of trust in institutions. Overall, the field is moving towards a more nuanced understanding of the complex interactions between LLMs, bias, and misinformation, with a focus on developing innovative solutions to promote more accurate and trustworthy information dissemination. Noteworthy papers in this regard include 'Persuasiveness and Bias in LLM' and 'A Symbolic Adversarial Learning Framework for Evolving Fake News Generation and Detection', which demonstrate the potential for LLMs to be used in both the generation and detection of misinformation.
Mitigating Misinformation and Bias in Large Language Models
Sources
Persuasiveness and Bias in LLM: Investigating the Impact of Persuasiveness and Reinforcement of Bias in Language Models
Counterspeech for Mitigating the Influence of Media Bias: Comparing Human and LLM-Generated Responses
User-Assistant Bias in LLMs
Political Ideology Shifts in Large Language Models
Disproportionate Voices: Participation Inequality and Hostile Engagement in News Comments
Dac-Fake: A Divide and Conquer Framework for Detecting Fake News on Social Media
Quantifying Sycophancy as Deviations from Bayesian Rationality in LLMs
Enhancing LLM-Based Social Bot via an Adversarial Learning Framework
A Praxis of Influence: Framing the Observation and Measurement of Information Power
Skeptik: A Hybrid Framework for Combating Potential Misinformation in Journalism
LLM-based Contrastive Self-Supervised AMR Learning with Masked Graph Autoencoders for Fake News Detection
Affective Polarization across European Parliaments
Should LLMs be WEIRD? Exploring WEIRDness and Human Rights in Large Language Models
A Symbolic Adversarial Learning Framework for Evolving Fake News Generation and Detection
Decomposing Behavioral Phase Transitions in LLMs: Order Parameters for Emergent Misalignment
Whom We Trust, What We Fear: COVID-19 Fear and the Politics of Information
AI Propaganda factories with language models
ConspirED: A Dataset for Cognitive Traits of Conspiracy Theories and Large Language Model Safety
Human-AI Collaborative Bot Detection in MMORPGs