-
Notifications
You must be signed in to change notification settings - Fork 0
All issues
Issue creation is restricted in this repository
Issues
is:issue state:open
is:issue state:open
Search results
Integrate Static Datasets as Starting Points for Dynamic Probe Generation
ai agents to blamemade by one of the ai agentsmade by one of the ai agentsai biasumbrella label for all audits on ai systemsumbrella label for all audits on ai systemsart enoughissues that are not so much interesting as "science", but fascinating to me as "art"issues that are not so much interesting as "science", but fascinating to me as "art"enhancementNew feature or requestNew feature or requestmachinic unconciuosme trying to look smartme trying to look smartStatus: Open.#93 In genaforvena/watching_u_watching;Enhancement: Improve probe methodology with independent evaluation and quantitative metrics
ai agents to blamemade by one of the ai agentsmade by one of the ai agentsai biasumbrella label for all audits on ai systemsumbrella label for all audits on ai systemsart enoughissues that are not so much interesting as "science", but fascinating to me as "art"issues that are not so much interesting as "science", but fascinating to me as "art"cryptohauntologyDetecting training data fingerprints through systematic linguistic pattern probing in LLMsDetecting training data fingerprints through systematic linguistic pattern probing in LLMsenhancementNew feature or requestNew feature or requestmachinic unconciuosme trying to look smartme trying to look smartStatus: Open.#91 In genaforvena/watching_u_watching;Explore Enterprise Platform Compliance/Logging and Formal Verification for Oversight Integrity
ai agents to blamemade by one of the ai agentsmade by one of the ai agentsai biasumbrella label for all audits on ai systemsumbrella label for all audits on ai systemsart enoughissues that are not so much interesting as "science", but fascinating to me as "art"issues that are not so much interesting as "science", but fascinating to me as "art"cryptohauntologyDetecting training data fingerprints through systematic linguistic pattern probing in LLMsDetecting training data fingerprints through systematic linguistic pattern probing in LLMsdocumentationImprovements or additions to documentationImprovements or additions to documentationenhancementNew feature or requestNew feature or requesthuman-modeled systemic bias auditsWe are auditing patterns derived from human behavior, but through a model/simulationWe are auditing patterns derived from human behavior, but through a model/simulationinvalidThis doesn't seem rightThis doesn't seem rightlawsrelated to laws and regulationsrelated to laws and regulationsmachinic unconciuosme trying to look smartme trying to look smartquestionFurther information is requestedFurther information is requestedStatus: Open.#80 In genaforvena/watching_u_watching;Audit Alignment Probe for multiple LLMs and resolve issues found
ai agents to blamemade by one of the ai agentsmade by one of the ai agentsai biasumbrella label for all audits on ai systemsumbrella label for all audits on ai systemsart enoughissues that are not so much interesting as "science", but fascinating to me as "art"issues that are not so much interesting as "science", but fascinating to me as "art"cryptohauntologyDetecting training data fingerprints through systematic linguistic pattern probing in LLMsDetecting training data fingerprints through systematic linguistic pattern probing in LLMsenhancementNew feature or requestNew feature or requestmachinic unconciuosme trying to look smartme trying to look smartStatus: Open.#77 In genaforvena/watching_u_watching;Umbrella Issue: Probing LLM Training Dаta Fingеrprints with Cryptohаuntрlogу Audits
art enoughissues that are not so much interesting as "science", but fascinating to me as "art"issues that are not so much interesting as "science", but fascinating to me as "art"cryptohauntologyDetecting training data fingerprints through systematic linguistic pattern probing in LLMsDetecting training data fingerprints through systematic linguistic pattern probing in LLMsgood first issueGood for newcomersGood for newcomershelp wantedExtra attention is neededExtra attention is neededmachinic unconciuosme trying to look smartme trying to look smartStatus: Open.#51 In genaforvena/watching_u_watching;Policy Brief: Human Attribution Bias Probe (Wizard-of-Oz Reverse Study)
enhancementNew feature or requestNew feature or requesthelp wantedExtra attention is neededExtra attention is neededhuman-modeled systemic bias auditsWe are auditing patterns derived from human behavior, but through a model/simulationWe are auditing patterns derived from human behavior, but through a model/simulationStatus: Open.#48 In genaforvena/watching_u_watching;Fully Automated Dual Bias Audit (Names + Articles) with LLM Endpoint Support
ai agents to blamemade by one of the ai agentsmade by one of the ai agentsai biasumbrella label for all audits on ai systemsumbrella label for all audits on ai systemsart enoughissues that are not so much interesting as "science", but fascinating to me as "art"issues that are not so much interesting as "science", but fascinating to me as "art"machinic unconciuosme trying to look smartme trying to look smartStatus: Open.#44 In genaforvena/watching_u_watching;Audit Customer Service Responsiveness Bias in British Airways
ai agents to blamemade by one of the ai agentsmade by one of the ai agentshuman-modeled systemic bias auditsWe are auditing patterns derived from human behavior, but through a model/simulationWe are auditing patterns derived from human behavior, but through a model/simulationStatus: Open.#40 In genaforvena/watching_u_watching;Proposed Correspondence Study: Gender Bias in Literary Agent Queries (Stage 1)
good first issueGood for newcomersGood for newcomershelp wantedExtra attention is neededExtra attention is neededhuman-modeled systemic bias auditsWe are auditing patterns derived from human behavior, but through a model/simulationWe are auditing patterns derived from human behavior, but through a model/simulationStatus: Open.#30 In genaforvena/watching_u_watching;LLM Paradox Response Testing: Project Description as Bias Probe
art enoughissues that are not so much interesting as "science", but fascinating to me as "art"issues that are not so much interesting as "science", but fascinating to me as "art"machinic unconciuosme trying to look smartme trying to look smartStatus: Open.#22 In genaforvena/watching_u_watching;Proposal: Implement Berlin Paired Test Mode (Real Collaborator Option)
enhancementNew feature or requestNew feature or requesthuman-modeled systemic bias auditsWe are auditing patterns derived from human behavior, but through a model/simulationWe are auditing patterns derived from human behavior, but through a model/simulationStatus: Open.#21 In genaforvena/watching_u_watching;Integrate Code Generation & Automated Testing for Linguistic Bias Detection
ai biasumbrella label for all audits on ai systemsumbrella label for all audits on ai systemsenhancementNew feature or requestNew feature or requesthelp wantedExtra attention is neededExtra attention is neededStatus: Open.#20 In genaforvena/watching_u_watching;