Channel Avatar

Tunadorable @UCeQhm8DwHBg_YEYY0KGM1GQ@youtube.com

7.47K subscribers - no pronouns :c

I am become change, confuser of subscribers


the dangers of centralized #ai risk of #ai causing #deflation #economics #inflation how does GPT-4o sound so natural? #ai #chatgpt #gpt4 #openai the effect of #ai on #ww3 quadratic vs sub-quadratic architectures. fyi the human brain is sub-quadratic #ai #transformer what happens after AGI replaces jobs? #ai #agi #job #economy sequentially dependent draft heads #ai #speculativedecoding https://arxiv.org/pdf/2402.05109.pdf do LLMs think like humans? what’s the definition of intelligence? #ai #intelligence https://arxiv.org/pdf/2312.09546v1.pdf do language models see visual illusions? https://arxiv.org/pdf/2311.00047.pdf #ai #illusion chatGPT takes its own notes #ai #arxiv https://arxiv.org/pdf/2305.00833.pdf transformers are multi-state RNNs #ai #transformers https://arxiv.org/pdf/2401.06104.pdf auto-parallel auto-regressive decoding #ai https://arxiv.org/pdf/2401.06761.pdf neurosymbolic AI is silly #ai #neurosymbolicai https://arxiv.org/pdf/2401.01040.pdf length extrapolation #ai #transformers #arxiv https://arxiv.org/abs/2312.17044 invitation to reinforcement learning #ai #reinforcementlearning https://arxiv.org/pdf/2312.08365.pdf https://www.lesswrong.com/posts/GpSzShaaf8po4rcmA/qapr-5-grokking-is-maybe-not-that-big-a-deal why grokking discovery was crazy #machinelearning #grok https://arxiv.org/pdf/2201.02177.pdf Emergent Abilities of LLMs #ai #emergence https://arxiv.org/pdf/2206.07682.pdf A Tale of Two Circuits #ai #machinelearning #grok #sparsity https://arxiv.org/pdf/2303.11873.pdf unifying grokking & double descent #ai #grok #doubledescent https://arxiv.org/pdf/2303.06173.pdf memorization in neural nets #machinelearning #ai #memorization https://arxiv.org/pdf/1706.05394.pdf towards understanding grokking #machinelearning #ai #grok https://arxiv.org/pdf/2205.10343.pdf grokking IS compression #machinelearning #ai #grok https://arxiv.org/pdf/2310.05918.pdf #machinelearning #ai #arxiv #grok https://arxiv.org/pdf/2309.02390.pdf probably safe systems #agi #singularity #apocalypse https://arxiv.org/pdf/2309.01933.pdf etymology of grokking #grok #ai #etymology grokking is not random #grok #machinelearning #ai #artificialintelligence #arxiv #machinelearning #ai #artificialintelligence https://arxiv.org/pdf/2310.03789.pdf demystifying embedding spaces using LLMs #machinelearning #ai https://arxiv.org/pdf/2310.04475.pdf over-simplification but a good eli5 #grok #machinelearning #ai grokking ‘grokking’ #ai #machinelearning #grok https://www.beren.io/2022-01-11-Grokking-Grokking/ #ai #machinelearning #arxiv #grok https://arxiv.org/pdf/2310.06110.pdf transformers solve mazes #ai #machinelearning #arxiv https://arxiv.org/pdf/2312.02566.pdf minimum hyperspherical energy #ai #machinelearning #arxiv https://arxiv.org/abs/1805.09298 sharpened cosine similarity #ai #machinelearning #arxiv https://arxiv.org/pdf/2307.13855.pdf deep hyperspherical learning #ai #machinelearning #arxiv https://arxiv.org/pdf/1711.03189.pdf hyperspherical prototype networks #ai #machinelearning #arxiv https://arxiv.org/pdf/1901.10514.pdf #ai #machinelearning https://openreview.net/attachment?id=OXourTLd9UO&name=supplementary_material cosine normalization #ai #machinelearning #arxiv https://arxiv.org/pdf/1702.05870.pdf Guide LM Reasoning with Plan Tokens #ai #machinelearning #arxiv https://arxiv.org/pdf/2310.05707.pdf #ai #machinelearning #arxiv https://arxiv.org/pdf/2305.18741.pdf #ai #machinelearning #artificialintelligence https://arxiv.org/pdf/2307.06945.pdf #machinelearning #ai #artificialintelligence https://arxiv.org/pdf/2310.01732.pdf