Aman Chadha
2 indexed papers
Recent (6 mo)
2With code
0Influential cites
0Benchmarked
0Publications per year
226
Top categories
NLP×2ML×2AI×1
Frequent co-authors
Research Timeline
2026
MAAT: Multi-phase Adapter-Aware Targeted Unlearning
The paper introduces 5WBENCH, a new benchmark for causal unlearning, and proposes MAAT, a novel three-phase framework that achieves high forgetting and high retention specifically on complex 'Why'-type causal knowledge.
MENTIS: What Belief Changes Under Alignment? Measuring Multi-Scale Latent Torsion in Language Models
The paper introduces MENTIS, a geometry-first framework that measures how preference alignment structurally changes the internal computations of language models, finding that these changes are selective, depth-localized, and concept-dependent.
Highlighted terms show continued research focus across papers
Papers
cs.CLcs.AIcs.LGRecentMay 31, 2026
MENTIS: What Belief Changes Under Alignment? Measuring Multi-Scale Latent Torsion in Language Models
The paper introduces MENTIS, a geometry-first framework that measures how preference alignment structurally changes the internal computations of language models, finding that these changes are selecti…
View →
cs.LGcs.CLRecentMay 28, 2026