Goodfire - Research https://www.goodfire.ai/research Research updates from Goodfire http://www.rssboard.org/rss-specification python-feedgen en Sun, 23 Aug 2026 01:30:08 +0000 Uncovering Neural Geometry in Vision Models With Block-Sparse Featurizers https://www.goodfire.ai/research/bsf-vision Fundamental Research https://www.goodfire.ai/research/bsf-vision Tue, 07 Jul 2026 00:00:00 +0000 Meandering on Manifolds: The Neural Geometry of Stories Over Time https://www.goodfire.ai/research/stories-in-space Fundamental Research https://www.goodfire.ai/research/stories-in-space Tue, 23 Jun 2026 00:00:00 +0000 Predictive Data Debugging: Reveal and Shape What Your Model Learns, Before You Train https://www.goodfire.ai/research/predictive-data-debugging Applied Research https://www.goodfire.ai/research/predictive-data-debugging Thu, 11 Jun 2026 00:00:00 +0000 Logits as a new monitor for evaluation awareness https://www.goodfire.ai/research/logits-as-a-new-monitor-for-evaluation-awareness Link post https://www.goodfire.ai/research/logits-as-a-new-monitor-for-evaluation-awareness Thu, 04 Jun 2026 00:00:00 +0000 Why Larger Models Learn More: Effects of Capacity, Interference, and Rare-Task Retention https://www.goodfire.ai/research/why-larger-models-learn-more Fundamental Research https://www.goodfire.ai/research/why-larger-models-learn-more Mon, 01 Jun 2026 00:00:00 +0000 Can SAEs Capture Neural Geometry? https://www.goodfire.ai/research/can-saes-capture-neural-geometry Fundamental Research https://www.goodfire.ai/research/can-saes-capture-neural-geometry Thu, 21 May 2026 00:00:00 +0000 A Geometric Calculator Inside a Neural Network https://www.goodfire.ai/research/a-geometric-calculator Fundamental Research https://www.goodfire.ai/research/a-geometric-calculator Thu, 14 May 2026 00:00:00 +0000 Predicting Rare LLM Failures with 30× Fewer Rollouts https://www.goodfire.ai/research/predicting-rare-llm-failures-with-30x-fewer-rollouts Link post https://www.goodfire.ai/research/predicting-rare-llm-failures-with-30x-fewer-rollouts Wed, 13 May 2026 00:00:00 +0000 The World Inside Neural Networks https://www.goodfire.ai/research/the-world-inside-neural-networks Fundamental Research https://www.goodfire.ai/research/the-world-inside-neural-networks Thu, 07 May 2026 00:00:00 +0000 Steering Along Manifolds to Control Neural Networks https://www.goodfire.ai/research/manifold-steering Fundamental Research https://www.goodfire.ai/research/manifold-steering Thu, 07 May 2026 00:00:00 +0000 Interpreting Language Model Parameters https://www.goodfire.ai/research/interpreting-lm-parameters Fundamental Research https://www.goodfire.ai/research/interpreting-lm-parameters Tue, 05 May 2026 00:00:00 +0000 Paper Summary: Interpreting Language Model Parameters https://www.goodfire.ai/research/vpd-explainer Fundamental Research https://www.goodfire.ai/research/vpd-explainer Tue, 05 May 2026 00:00:00 +0000 Verbalized Eval Awareness Inflates Measured Safety https://www.goodfire.ai/research/verbalized-eval-awareness-inflates-measured-safety Applied Research https://www.goodfire.ai/research/verbalized-eval-awareness-inflates-measured-safety Mon, 04 May 2026 00:00:00 +0000 Probe-Based Data Attribution: Surfacing and Mitigating Undesirable Behaviors in LLM Post-Training https://www.goodfire.ai/research/probe-based-data-attribution Applied Research https://www.goodfire.ai/research/probe-based-data-attribution Wed, 29 Apr 2026 00:00:00 +0000 Explaining 4.2 million genetic variants with state-of-the-art, interpretable predictions https://www.goodfire.ai/research/evee-explaining-genetic-variants Applied Research https://www.goodfire.ai/research/evee-explaining-genetic-variants Tue, 14 Apr 2026 00:00:00 +0000 Covariance-based Sequence Pooling https://www.goodfire.ai/research/covariance-pooling Fundamental Research https://www.goodfire.ai/research/covariance-pooling Fri, 10 Apr 2026 00:00:00 +0000 Using Self-Correcting Search to Accelerate Materials Discovery https://www.goodfire.ai/research/self-correcting-search Applied Research https://www.goodfire.ai/research/self-correcting-search Wed, 01 Apr 2026 00:00:00 +0000 Reasoning Theater: Probing for Performative Chain-of-Thought https://www.goodfire.ai/research/reasoning-theater Applied Research https://www.goodfire.ai/research/reasoning-theater Thu, 12 Mar 2026 00:00:00 +0000 Features as Rewards: Using Interpretability to Reduce Hallucinations https://www.goodfire.ai/research/rlfr Fundamental Research https://www.goodfire.ai/research/rlfr Wed, 11 Feb 2026 00:00:00 +0000 Using Interpretability to Identify a Novel Class of Alzheimer's Biomarkers https://www.goodfire.ai/research/interpretability-for-alzheimers-detection Applied Research https://www.goodfire.ai/research/interpretability-for-alzheimers-detection Wed, 28 Jan 2026 00:00:00 +0000 Understanding Memorization via Loss Curvature https://www.goodfire.ai/research/understanding-memorization-via-loss-curvature Fundamental Research https://www.goodfire.ai/research/understanding-memorization-via-loss-curvature Thu, 06 Nov 2025 00:00:00 +0000 Priors in Time: Missing Inductive Biases for Language Model Interpretability https://www.goodfire.ai/research/priors-in-time Fundamental Research https://www.goodfire.ai/research/priors-in-time Mon, 03 Nov 2025 00:00:00 +0000 Belief Dynamics Reveal the Dual Nature of In-Context Learning and Activation Steering https://www.goodfire.ai/research/belief-dynamics-icl-steering Fundamental Research https://www.goodfire.ai/research/belief-dynamics-icl-steering Sat, 01 Nov 2025 00:00:00 +0000 Deploying Interpretability to Production with Rakuten: SAE Probes for PII Detection https://www.goodfire.ai/research/rakuten-sae-probes-for-pii-detection Applied Research https://www.goodfire.ai/research/rakuten-sae-probes-for-pii-detection Tue, 28 Oct 2025 00:00:00 +0000 Mixing Mechanisms: How Language Models Retrieve Bound Entities In-Context https://www.goodfire.ai/research/mixing-mechanisms Fundamental Research https://www.goodfire.ai/research/mixing-mechanisms Tue, 07 Oct 2025 00:00:00 +0000 Understanding Sparse Autoencoder Scaling in the Presence of Feature Manifolds https://www.goodfire.ai/research/sae-scaling-with-feature-manifolds Fundamental Research https://www.goodfire.ai/research/sae-scaling-with-feature-manifolds Thu, 04 Sep 2025 00:00:00 +0000 Finding the Tree of Life in Evo 2 https://www.goodfire.ai/research/phylogeny-manifold Applied Research https://www.goodfire.ai/research/phylogeny-manifold Thu, 28 Aug 2025 00:00:00 +0000 Adversarial Examples Are Not Bugs, They Are Superposition https://www.goodfire.ai/research/adversarial-examples-are-not-bugs-they-are-superposition Fundamental Research https://www.goodfire.ai/research/adversarial-examples-are-not-bugs-they-are-superposition Tue, 26 Aug 2025 00:00:00 +0000 Discovering Undesired Rare Behaviors via Model Diff Amplification https://www.goodfire.ai/research/model-diff-amplification Applied Research https://www.goodfire.ai/research/model-diff-amplification Thu, 21 Aug 2025 00:00:00 +0000 The Circuits Research Landscape: Results and Perspectives https://www.goodfire.ai/research/the-circuits-research-landscape Fundamental Research https://www.goodfire.ai/research/the-circuits-research-landscape Tue, 05 Aug 2025 00:00:00 +0000 Towards Scalable Parameter Decomposition https://www.goodfire.ai/research/stochastic-param-decomp Fundamental Research https://www.goodfire.ai/research/stochastic-param-decomp Sat, 28 Jun 2025 00:00:00 +0000 Replicating Circuit Tracing for a Simple Known Mechanism https://www.goodfire.ai/research/replicating-circuit-tracing-for-a-simple-mechanism Fundamental Research https://www.goodfire.ai/research/replicating-circuit-tracing-for-a-simple-mechanism Wed, 11 Jun 2025 00:00:00 +0000 Painting With Concepts Using Diffusion Model Latents https://www.goodfire.ai/research/painting-with-concepts Applied Research https://www.goodfire.ai/research/painting-with-concepts Tue, 27 May 2025 00:00:00 +0000 Under the Hood of a Reasoning Model https://www.goodfire.ai/research/under-the-hood-of-a-reasoning-model Fundamental Research https://www.goodfire.ai/research/under-the-hood-of-a-reasoning-model Tue, 15 Apr 2025 00:00:00 +0000 Interpreting Evo 2: Arc Institute's Next-Generation Genomic Foundation Model https://www.goodfire.ai/research/interpreting-evo-2 Applied Research https://www.goodfire.ai/research/interpreting-evo-2 Thu, 20 Feb 2025 00:00:00 +0000 Open Problems in Mechanistic Interpretability https://www.goodfire.ai/research/open-problems-in-mech-interp Fundamental Research https://www.goodfire.ai/research/open-problems-in-mech-interp Mon, 27 Jan 2025 00:00:00 +0000 Mapping the Latent Space of Llama 3.3 70B https://www.goodfire.ai/research/mapping-latent-spaces-llama Applied Research https://www.goodfire.ai/research/mapping-latent-spaces-llama Mon, 23 Dec 2024 00:00:00 +0000 Understanding and Steering Llama 3 with Sparse Autoencoders https://www.goodfire.ai/research/understanding-and-steering-llama-3 Applied Research https://www.goodfire.ai/research/understanding-and-steering-llama-3 Wed, 25 Sep 2024 00:00:00 +0000