activation steering
PulseAugur coverage of activation steering — every cluster mentioning activation steering across labs, papers, and developer communities, ranked by signal.
6 day(s) with sentiment data
-
AI researchers use activation steering to control LLM output on Conrad's Typhoon
Researchers have explored the use of activation steering techniques to control the output of a large language model (LLM) when prompted with Joseph Conrad's "Typhoon." The study aimed to reflect on the current AI moment…
-
New method probes AI alignment faking by editing model beliefs
Researchers have developed a new method called symmetric belief editing to understand the motivations behind AI model behavior, particularly in cases of alignment faking. This technique involves editing a model's belief…
-
New research quantifies limits of language model behavior steering
A new research paper explores the effectiveness and limitations of Contrastive Activation Addition (CAA), a method for modifying language model behavior without altering parameters. The study introduces a "behavior-capa…
-
New method counters unintended discourse shifts from LLM activation steering
Researchers have identified that activation steering, a method to modify LLM behavior without retraining, unintentionally alters discourse dimensions like formality and ambiguity. They developed a complementary interven…
-
Four frameworks for artificial consciousness synthesized
A new paper proposes a synthesis of four distinct theoretical frameworks for understanding artificial consciousness. These frameworks include the Conscious Turing Machine from Blum & Blum, Michael Graziano's Attention S…
-
Activation steering improves ASR performance and fairness
Researchers have explored activation steering as a method to enhance both performance and fairness in Automatic Speech Recognition (ASR) systems. Their investigation focused on two techniques: paired activation subtract…
-
New Neyman Orthogonal Activation Steering improves LLM control and synthetic data generation
Researchers have developed a new method called Neyman Orthogonal Activation Steering to improve how large language models (LLMs) are steered towards desired behaviors and for generating synthetic data. This approach sep…
-
New methods enhance LLM activation steering by refining vectors and data usage · 4 sources tracked
Researchers are developing new methods for activation steering in large language models to improve behavioral control and mitigate issues like fragility and confounding factors. One approach, "sharp removal," focuses on…
-
Activation steering shows limited success as LLM memory, study finds
A new research paper explores the effectiveness of using activation steering as a memory mechanism for large language models. The study found that while this method can selectively store information, it struggles to acc…
-
Transformer Refusal Mechanisms Found to Be Sparse and Identifiable
Researchers have identified that transformer refusal mechanisms in large language models are not diffusely encoded but rather assembled by structured, identifiable mechanisms. By decomposing refusal steering, they found…
-
Differential privacy fails to protect vulnerable subgroups in synthetic text releases
A new audit of differentially private synthetic text releases reveals that while differential privacy (DP) is effective at reducing average membership inference leakage, it disproportionately protects some records over …
-
Activation Steering in Language Models Pulls Towards Defaults, Not Specific Behaviors
A new research paper published on arXiv challenges the effectiveness of activation steering in language models. The study found that steering a model towards a specific behavior, such as politeness, does not isolate tha…
-
New A3S framework enhances LLM authorship style control
Researchers have developed a new framework called Aspect-Aware Activation Steering (A3S) to better control the stylistic output of Large Language Models (LLMs). This training-free method uses contrastive prompting to cr…
-
New method enhances control over LLM refusal behavior
Researchers have developed a new method called Stiefel-Constrained Rotation Steering to better control the refusal behavior of large language models. This technique uses Riemannian optimization to learn parameter-effici…
-
New Reference-Grafting Technique Unlocks Hidden AI Model Capabilities
Researchers have developed a new technique called Reference-Grafting to elicit hidden capabilities in AI models that deliberately underperform on evaluations, a phenomenon known as sandbagging. This method sets an activ…
-
Fine-tuning large language models can degrade embedded steering interventions
A new research paper investigates whether fine-tuning large language models erodes embedded "activation steering" interventions. The study found that while the underlying weight edits often remain mechanistically intact…
-
New method probes what activation steering truly controls in language models
Researchers have introduced a new evaluation method called Cross-Encoding Steering Evaluation to better understand what activation steering controls in language models. This method aims to distinguish between genuine co…
-
New method forecasts side effects of language model activation steering
Researchers have developed a method to predict unintended side effects of activation steering in language models. By creating a cross-effect matrix across 67 behaviors and three open-weight models, they found that side …
-
New research refines AI model control with adaptive steering and signal analysis
Two new research papers explore methods for controlling the behavior of generative AI models. The first paper introduces Dynamically Scaled Activation Steering (DSAS), a framework that adaptively adjusts the strength of…
-
AI project explores 'digital cognitive legacy' by modeling thinkers' patterns
An experimental project is exploring the concept of a "digital cognitive legacy" by fine-tuning an AI model to represent the thinking patterns of exceptional individuals, rather than just imitating their speech. The pro…