6 papers
ResidualQuant: KV Cache Quantization for Looped Transformers with 2-Bit Residuals
Heejun Kim, Junyoung Lee, SangLyul Cho +3
Looped Transformers improve parameter efficiency by applying shared Transformer blocks over recurrent loops, increasing computational depth without increasing the parameter count.…
Event-by-event net-charge fluctuations in pp, OO, Ne-Ne, and Pb-Pb collisions with ALICE at the LHC
Nida Malik
Investigations involving the event-by-event fluctuations of conserved quantities, like net charge, net baryon number, and strangeness in heavy-ion collisions, provide insights into…
LoopSpec: Pipelined Self-Speculative Decoding for Looped Transformers
SangLyul Cho, Langqing Cui, Sehoon Kim +2
Looped Transformers achieve strong performance with compact parameter sizes by repeatedly applying a shared stack of Transformer blocks across recurrent depths. However, they incur…
Pramana: A Composable, Domain-Specific Backend for Empirical Networking Research
Jaber Daneshamooz, Eugene Vuong, Alagappan Ramanathan +11
Networking research advances by turning hypotheses into empirical evidence, so accelerating it means reducing the lag between ideation (synthesizing a hypothesis) and generating th…
Towards an Agentic Workflow for Internet Measurement Research
Alagappan Ramanathan, Eunju Kang, Dongsu Han +1
Internet measurement research faces an accessibility crisis: complex analyses require custom integration of multiple specialized tools that demands specialized domain expertise. Wh…
SpecEdge: Scalable Edge-Assisted Serving Framework for Interactive LLMs
Jinwoo Park, Seunggeun Cho, Dongsu Han
Large language models (LLMs) power many modern applications, but serving them at scale remains costly and resource-intensive. Current server-centric systems overlook consumer-grade…