https://www.goodfire.com/https://www.goodfire.com/bloghttps://www.goodfire.com/blog/a-practical-guide-to-saeshttps://www.goodfire.com/blog/ai-safety-still-needs-great-engineershttps://www.goodfire.com/blog/announcing-goodfire-research-grantshttps://www.goodfire.com/blog/announcing-our-50m-series-ahttps://www.goodfire.com/blog/ap293-guest-lectures-25https://www.goodfire.com/blog/feature-steering-for-reliable-and-expressive-ai-engineeringhttps://www.goodfire.com/blog/fellowship-fall-25https://www.goodfire.com/blog/goodfire-baseten-partnershiphttps://www.goodfire.com/blog/intentional-designhttps://www.goodfire.com/blog/interpretability-infra-at-frontier-scalehttps://www.goodfire.com/blog/mayo-clinic-collaborationhttps://www.goodfire.com/blog/on-optimism-for-interpretabilityhttps://www.goodfire.com/blog/our-approach-to-safetyhttps://www.goodfire.com/blog/our-series-bhttps://www.goodfire.com/blog/probe-monitors-101https://www.goodfire.com/blog/radical-partnership-announcementhttps://www.goodfire.com/blog/sae-open-source-announcementhttps://www.goodfire.com/blog/soc-2-type-iihttps://www.goodfire.com/blog/we-can-and-must-solve-alignmenthttps://www.goodfire.com/blog/you-and-your-research-agenthttps://www.goodfire.com/careershttps://www.goodfire.com/companyhttps://www.goodfire.com/contact-ushttps://www.goodfire.com/customer-stories/prima-mentehttps://www.goodfire.com/customer-stories/rakutenhttps://www.goodfire.com/demos/dolci-viewerhttps://www.goodfire.com/demos/hallucinations-viewerhttps://www.goodfire.com/languagehttps://www.goodfire.com/legal/msahttps://www.goodfire.com/legal/pilot-agreementhttps://www.goodfire.com/legal/privacyhttps://www.goodfire.com/legal/toshttps://www.goodfire.com/legal/websitetoshttps://www.goodfire.com/life-scienceshttps://www.goodfire.com/request-demohttps://www.goodfire.com/researchhttps://www.goodfire.com/research/a-geometric-calculatorhttps://www.goodfire.com/research/already-cleanhttps://www.goodfire.com/research/belief-dynamics-icl-steeringhttps://www.goodfire.com/research/better-biosecurity-monitorshttps://www.goodfire.com/research/bsf-visionhttps://www.goodfire.com/research/can-saes-capture-neural-geometryhttps://www.goodfire.com/research/covariance-poolinghttps://www.goodfire.com/research/evee-explaining-genetic-variantshttps://www.goodfire.com/research/forking-fasthttps://www.goodfire.com/research/interpretability-for-alzheimers-detectionhttps://www.goodfire.com/research/interpreting-evo-2https://www.goodfire.com/research/interpreting-lm-parametershttps://www.goodfire.com/research/logits-as-a-new-monitor-for-evaluation-awarenesshttps://www.goodfire.com/research/manifold-steeringhttps://www.goodfire.com/research/mapping-latent-spaces-llamahttps://www.goodfire.com/research/mixing-mechanismshttps://www.goodfire.com/research/model-diff-amplificationhttps://www.goodfire.com/research/neural-geometryhttps://www.goodfire.com/research/open-problems-in-mech-interphttps://www.goodfire.com/research/painting-with-conceptshttps://www.goodfire.com/research/phylogeny-manifoldhttps://www.goodfire.com/research/predicting-rare-llm-failures-with-30x-fewer-rolloutshttps://www.goodfire.com/research/predictive-data-debugginghttps://www.goodfire.com/research/priors-in-timehttps://www.goodfire.com/research/probe-based-data-attributionhttps://www.goodfire.com/research/production-cyber-monitors-on-kimi-k3https://www.goodfire.com/research/rakuten-sae-probes-for-pii-detectionhttps://www.goodfire.com/research/reasoning-theaterhttps://www.goodfire.com/research/replicating-circuit-tracing-for-a-simple-mechanismhttps://www.goodfire.com/research/reward-hacking-activation-monitorshttps://www.goodfire.com/research/rlfrhttps://www.goodfire.com/research/sae-scaling-with-feature-manifoldshttps://www.goodfire.com/research/self-correcting-searchhttps://www.goodfire.com/research/stochastic-param-decomphttps://www.goodfire.com/research/stories-in-spacehttps://www.goodfire.com/research/the-world-inside-neural-networkshttps://www.goodfire.com/research/under-the-hood-of-a-reasoning-modelhttps://www.goodfire.com/research/understanding-and-steering-llama-3https://www.goodfire.com/research/understanding-memorization-via-loss-curvaturehttps://www.goodfire.com/research/verbalized-eval-awareness-inflates-measured-safetyhttps://www.goodfire.com/research/vpd-explainerhttps://www.goodfire.com/research/why-larger-models-learn-more