IBM (United States)

Neuralign places neural recordings from different sessions and animals, with different sampled neurons, into one tensor decomposition in which each component shares a single temporal curve while trial strength and neuron weights are estimated per session, so no matched neurons or matched trials are needed. On 93 brain-wide recordings from 19 mice, an unsupervised fit yielded a single component that separated the three anterior lateral motor cortex (ALM) optogenetic inhibition conditions, with balanced accuracy of 0.647 on held-out sessions against a chance level of 1/3. That component's change showed a modest association with the drop in behavioural accuracy (ρ = 0.266 over 180 session–inhibition observations).

Neuralign places neural recordings from different sessions and animals, with different sampled neurons, into one tensor decomposition in which each component shares a single temporal curve while trial strength and neuron weights are estimated per session, so no matched neurons or matched trials are needed. On 93 brain-wide recordings from 19 mice, an unsupervised fit yielded a single component that separated the three anterior lateral motor cortex (ALM) optogenetic inhibition conditions, with balanced accuracy of 0.647 on held-out sessions against a chance level of 1/3. That component's change showed a modest association with the drop in behavioural accuracy (ρ = 0.266 over 180 session–inhibition observations).

AlgoPilot splits language-model problem solving into three steps: decide whether a problem calls for natural-language reasoning, code execution, constraint solving, symbolic planning or mathematical optimization; generate a structured formulation for the chosen paradigm; then let an execution model solve it. With Qwen3-8B as the executor, average accuracy on 13 benchmarks spanning 66 domains rises from 38.4% under direct prompting to 72.9%, 11.7 points above the best large-model prompting baseline GLM-5; with Gemma-4-E4B-it as the executor, average success rises from 39.0% under direct prompting to 60.5%.

AlgoPilot splits language-model problem solving into three steps: decide whether a problem calls for natural-language reasoning, code execution, constraint solving, symbolic planning or mathematical optimization; generate a structured formulation for the chosen paradigm; then let an execution model solve it. With Qwen3-8B as the executor, average accuracy on 13 benchmarks spanning 66 domains rises from 38.4% under direct prompting to 72.9%, 11.7 points above the best large-model prompting baseline GLM-5; with Gemma-4-E4B-it as the executor, average success rises from 39.0% under direct prompting to 60.5%.

BLIS is a deterministic, CPU-only discrete-event simulator that models the whole LLM inference serving pipeline, including admission, routing, scheduling, prefill/decode execution, KV-cache management, and preemption, reaching 6.7% median end-to-end error across 36 experiments while running about 200 times faster than real execution.

BLIS is a deterministic, CPU-only discrete-event simulator that models the whole LLM inference serving pipeline, including admission, routing, scheduling, prefill/decode execution, KV-cache management, and preemption, reaching 6.7% median end-to-end error across 36 experiments while running about 200 times faster than real execution.

The work introduces Universal Attention, which combines query-aware, key-aware, and similarity-based decay mechanisms via a geometric mean into an additive bias on the attention matrix, letting the model adaptively prune its KV-cache while keeping RoPE and Softmax; on a 1B Llama3-style model trained on 50B tokens it achieves over 10x compression, with RULER averages of 50.3/49.3 versus 40.8 for unpruned Llama3 at only 21.4/12.9 MB of cache.

The work introduces Universal Attention, which combines query-aware, key-aware, and similarity-based decay mechanisms via a geometric mean into an additive bias on the attention matrix, letting the model adaptively prune its KV-cache while keeping RoPE and Softmax; on a 1B Llama3-style model trained on 50B tokens it achieves over 10x compression, with RULER averages of 50.3/49.3 versus 40.8 for unpruned Llama3 at only 21.4/12.9 MB of cache.

The work introduces a unified formulation of context optimization that interprets an agent memory system update as an optimization update procedure over the model's context, and on that basis proposes GraphMemory, a lightweight graph-based memory that accumulates, refines, organizes, and connects reusable strategies; each query retrieves only the relevant subgraph, so under bounded retrieval the amount of retrieved memory stays constant as the number of processed examples grows, and experiments show competitive downstream performance while using approximately 81-85% fewer memory-construction tokens than the baselines.

The work introduces a unified formulation of context optimization that interprets an agent memory system update as an optimization update procedure over the model's context, and on that basis proposes GraphMemory, a lightweight graph-based memory that accumulates, refines, organizes, and connects reusable strategies; each query retrieves only the relevant subgraph, so under bounded retrieval the amount of retrieved memory stays constant as the number of processed examples grows, and experiments show competitive downstream performance while using approximately 81-85% fewer memory-construction tokens than the baselines.

The authors present FluidPD, a prefill-decode (P/D) disaggregated LLM serving system built on SGLang that offloads a bounded prefill suffix to decode workers with slack via FluidToken and switches running workers between prefill and decode roles in place via FluidRole without reloading models, improving overall SLO attainment over static SGLang by up to 94.6 percentage points on Azure-trace-derived staged and mixed workloads and by up to 88.3 percentage points over Dynamo autoscaling while using fewer GPUs.

The authors present FluidPD, a prefill-decode (P/D) disaggregated LLM serving system built on SGLang that offloads a bounded prefill suffix to decode workers with slack via FluidToken and switches running workers between prefill and decode roles in place via FluidRole without reloading models, improving overall SLO attainment over static SGLang by up to 94.6 percentage points on Azure-trace-derived staged and mixed workloads and by up to 88.3 percentage points over Dynamo autoscaling while using fewer GPUs.

The work proposes SHarP, which treats tools, instructions, and supporting mechanisms in an agent harness as individually disableable modules, estimates each module's performance and efficiency saliency through single-module ablation, and prunes by an efficiency-oriented or performance-oriented rule; across five settings (OpenHands/GAIA, PaperQA2/LitQA2, LIFE-harness airline and retail, JIT agentfold/DeepPlanning-Shopping), most pruned configurations retain comparable performance, all four harnesses achieve at least a 10% token-cost reduction, and two improve task performance while cutting cost.

The work proposes SHarP, which treats tools, instructions, and supporting mechanisms in an agent harness as individually disableable modules, estimates each module's performance and efficiency saliency through single-module ablation, and prunes by an efficiency-oriented or performance-oriented rule; across five settings (OpenHands/GAIA, PaperQA2/LitQA2, LIFE-harness airline and retail, JIT agentfold/DeepPlanning-Shopping), most pruned configurations retain comparable performance, all four harnesses achieve at least a 10% token-cost reduction, and two improve task performance while cutting cost.

The authors propose SWIFT, a framework that asynchronously co-serves text generation and watermark construction on the same vLLM backend, LLM, and GPU: it uses instruction-guided candidate generation with entity protection to produce context-aware substitutions, embeds watermarks via key-conditioned Tournament selection, and adapts watermark request admission with a scheduler driven by queue length and waiting-time pressure; on C4 text generation it attains the highest utility score of 4.87, 99.65% detection accuracy, and 0.905 s latency, retains 97.7% detection under substitution attacks, preserves task accuracy on GSM8K and MedMCQA, and reaches 528.7 tokens/s throughput with a 96.8% prefix-cache hit rate.

The authors propose SWIFT, a framework that asynchronously co-serves text generation and watermark construction on the same vLLM backend, LLM, and GPU: it uses instruction-guided candidate generation with entity protection to produce context-aware substitutions, embeds watermarks via key-conditioned Tournament selection, and adapts watermark request admission with a scheduler driven by queue length and waiting-time pressure; on C4 text generation it attains the highest utility score of 4.87, 99.65% detection accuracy, and 0.905 s latency, retains 97.7% detection under substitution attacks, preserves task accuracy on GSM8K and MedMCQA, and reaches 528.7 tokens/s throughput with a 96.8% prefix-cache hit rate.

The work proposes MaskCoFT, a masked co-adaptive fine-tuning method that trains routers and experts together using only the cross-entropy loss: during fine-tuning a learnable binary mask restricts each layer's Top-K routing to a subset of experts so the experts adapt to the redirected tokens, and at inference the learned mask becomes a soft prior that re-ranks experts while every expert remains selectable; with a simulated GPU cache of 4 experts per layer for Mixtral-8x7B and 12 for DeepSeek-V2-Lite, expert fetches per token drop by 23.7% and 10.1% relative to the base model, time per output token in real offloading system serving falls by up to 16.4% and 5.5%, and average accuracy over nine benchmarks stays above the base model by 0.92 and 0.53 points.

The work proposes MaskCoFT, a masked co-adaptive fine-tuning method that trains routers and experts together using only the cross-entropy loss: during fine-tuning a learnable binary mask restricts each layer's Top-K routing to a subset of experts so the experts adapt to the redirected tokens, and at inference the learned mask becomes a soft prior that re-ranks experts while every expert remains selectable; with a simulated GPU cache of 4 experts per layer for Mixtral-8x7B and 12 for DeepSeek-V2-Lite, expert fetches per token drop by 23.7% and 10.1% relative to the base model, time per output token in real offloading system serving falls by up to 16.4% and 5.5%, and average accuracy over nine benchmarks stays above the base model by 0.92 and 0.53 points.