The Amplifying Mirror: Locating and Steering the Partisan Direction inside a Large Language Model

• URL: http://arxiv.org/abs/2606.08792v1
• Date: Sun, 07 Jun 2026 19:17:02 GMT
• Title: The Amplifying Mirror: Locating and Steering the Partisan Direction inside a Large Language Model
• Abstract summary: We show that partisan political identity is encoded in the model's activation space.<n>Using 190,491 tweets from sitting members of the U.S. Congress as labeled training data, we train linear probes on the hidden states of the Llama 3.1 8B Instruct model.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large language models are rapicly replacing search engines as the primary interface between people and information. Unlike search engines, which retrieve existing content, LLMs generate novel text shaped by internal representations learned during training. Here we show that partisan political identity is encoded in the model's activation space, and that this direction directly shapes generation. Using 190,491 tweets from sitting members of the U.S. Congress as labeled training data, we train linear probes on the hidden states of the Llama 3.1 8B Instruct model. We identify a single geometric axis at layer 18 that separates Republican from Democratic text with an AUC of 0.945 and a Cohen's d of 1.94, and use sparse autoencoders to decompose that axis into interpretable partisan features. Causally intervening along this axis, ablating or amplifying the partisan component mid-generation, produces systematic shifts in the model's output. We witness stance reversals, register shifting, and structured fabrications of authority. Our results demonstrate that partisan bias in language models is not a vague emergent property but a learned geometric feature that can be precisely located and steered. Partisan bias is not a bug to be patched, but a structural property of how these models encode information about their users. As LLMs displace search engines as the interface to knowledge, understanding that product design (and its consequences) will be essential for navigating the legal, social, and political transitions from an information ecosystem that is curated to one that is generated.

Related papers

• From Echo Chambers to Epistemic Monoculture: Large Language Models Present Temporally Contingent Partisan Alignments as Knowledge [0.0]
  We present experiments that exploit a model's training cutoff in 2024.<n>We find that the model presents temporally contingent partisan alignments as knowledge, with no mechanism for distinguishing fact from opinion.
  arXiv  Detail & Related papers  (2026-09-07T16:41:56Z)
• The Neutral Mask: How RLHF Provides Shallow Alignment while Leaving Partisan Structure Intact in a Large Language Model [0.0]
  We show that RLHF does not remove the structured partisan direction in the base model.<n>We show that RLHF encodes a norm of political neutrality, not by erasing the model's knowledge of partisanship, but by severing the causal pathway from partisan geometry to output generation.
  arXiv  Detail & Related papers  (2026-06-08T17:00:31Z)
• A Unified Masked Jigsaw Puzzle Framework for Vision and Language Models [109.4033233070067]
  The gradient of Position Embeddings (PEs) in Transformer contains sufficient information, which can be used to reconstruct the input data.<n>We introduce a Masked Jigsaw Puzzle (MJP) framework to improve Transformer models' robustness against gradient attacks.<n>Results suggest that MJP is a unified framework for different Transformer-based models in both vision and language tasks.
  arXiv  Detail & Related papers  (2026-01-17T13:32:32Z)
• Finding the Translation Switch: Discovering and Exploiting the Task-Initiation Features in LLMs [69.28193153685893]
  Large Language Models (LLMs) frequently exhibit strong translation abilities, even without task-specific fine-tuning.<n>To demystify this process, we leverage Sparse Autoencoders (SAEs) and introduce a novel framework for identifying task-specific features.<n>Our work not only decodes a core component of the translation mechanism in LLMs but also provides a blueprint for using internal model mechanism to create more robust and efficient models.
  arXiv  Detail & Related papers  (2026-01-16T06:29:07Z)
• GPT and Prejudice: A Sparse Approach to Understanding Learned Representations in Large Language Models [0.0]
  Large language models (LLMs) are increasingly trained on massive, uncurated corpora.<n>We show that pairing LLMs with sparse autoencoders (SAEs) enables interpretation not only of model behavior but also of the deeper structures, themes, and biases embedded in the training data.<n>We train a GPT-style transformer model exclusively on the novels of Jane Austen, a corpus rich in social constructs and narrative patterns.
  arXiv  Detail & Related papers  (2025-09-24T11:10:16Z)
• Large Language Models Encode Semantics in Low-Dimensional Linear Subspaces [31.401762286885656]
  Understanding the latent space geometry of large language models (LLMs) is key to their behavior and alignment.<n>We conduct a large-scale study in 11 empirical models across 6 scientific topics.
  arXiv  Detail & Related papers  (2025-07-13T17:03:25Z)
• A Geometric Notion of Causal Probing [85.49839090913515]
  The linear subspace hypothesis states that, in a language model's representation space, all information about a concept such as verbal number is encoded in a linear subspace.<n>We give a set of intrinsic criteria which characterize an ideal linear concept subspace.<n>We find that, for at least one concept across two languages models, the concept subspace can be used to manipulate the concept value of the generated word with precision.
  arXiv  Detail & Related papers  (2023-07-27T17:57:57Z)
• Unsupervised Sentiment Analysis of Plastic Surgery Social Media Posts [91.3755431537592]
  The massive collection of user posts across social media platforms is primarily untapped for artificial intelligence (AI) use cases. Natural language processing (NLP) is a subfield of AI that leverages bodies of documents, known as corpora, to train computers in human-like language understanding. This study demonstrates that the applied results of unsupervised analysis allow a computer to predict either negative, positive, or neutral user sentiment towards plastic surgery.
  arXiv  Detail & Related papers  (2023-07-05T20:16:20Z)
• The Curse of Recursion: Training on Generated Data Makes Models Forget [70.02793975243212]
  Large language models (LLMs) are here to stay, and will bring about drastic change in the whole ecosystem of online text and images. We find that use of model-generated content in training causes irreversible defects in the resulting models, where tails of the original content distribution disappear.
  arXiv  Detail & Related papers  (2023-05-27T15:10:41Z)
• AAformer: Auto-Aligned Transformer for Person Re-Identification [82.45385078624301]
  We introduce an alignment scheme in transformer architecture for the first time. We propose the auto-aligned transformer (AAformer) to automatically locate both the human parts and nonhuman ones at patch level. AAformer integrates the part alignment into the self-attention and the output [PART]s can be directly used as part features for retrieval.
  arXiv  Detail & Related papers  (2021-04-02T08:00:25Z)
• Mind The Facts: Knowledge-Boosted Coherent Abstractive Text Summarization [47.47707316595568]
  We propose a novel architecture that extends Transformer encoder-decoder architecture. We incorporate entity-level knowledge from the Wikidata knowledge graph into the encoder-decoder architecture. This helps our model with producing coherent summaries even when the source article is long.
  arXiv  Detail & Related papers  (2020-06-27T20:06:15Z)

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.