ai coding agents are buddies, not replacements, says cognition ceo
cognition ceo scott wu says devin ai coding agent is designed to assist human programmers, not replace them, handling maintenance tasks to free up creative work.
ailatest summaries
plain 200-word summaries of important ai and data science news, updated every few hours.
deep learning for bivariate survival treatment rules
a new method uses deep neural networks and adaptive prediction-powered learning to optimize treatment rules for bivariate survival outcomes in randomized trials.
aigroq seeks $650m after nvidia deal
ai chip startup groq is reportedly raising $650 million from existing investors to expand its inference cloud business, following a $20 billion licensing and talent deal with nvidia.
aigoogle demos gemini omni and 3.5 flash capabilities
google shows nine demos of gemini omni for video creation and editing, and gemini 3.5 flash for agentic tasks and interactive web uis.
aillm reviews show limited alignment with human judgment
a study of acl rolling review papers finds llm reviews align only moderately with human reviews and can be gamed by authors using iterative revision.
aicategory theory boosts language model perplexity
a 306m-parameter transformer with simplicial message passing reduces perplexity by 12% over gpt-2 small on wikitext-103.
aiauditing job descriptions with textstat
use python's textstat library to automatically detect overly complex language in entry-level job postings.
airun nlp in the browser with transformers.js
a tutorial on using transformers.js for text classification, zero-shot labeling, and question answering directly in the browser with no server needed.
aibeginner's guide to pytorch profiling
learn to read torch.profiler traces and tables to find bottlenecks in pytorch code, starting with a simple matrix multiply and add.
aixcena raises $135m to put compute inside memory chips
startup xcena secures series b funding for a chip that moves data processing closer to dram, aiming to cut ai inference costs by reducing data movement between memory and processors.
aimulti-task ppi boosts inference with few labels
a new framework uses labeled data from related tasks to improve statistical power in prediction-powered inference when only a handful of labels are available per task.
aione mask to rule them all: hidden facts after editing
a compact binary mask reverses most knowledge edits in language models, revealing a shared mechanism behind diverse factual updates.
aibehavior-induced mirror-prox td for faster off-policy learning
a new mirror-prox temporal-difference method uses behavior-policy transition information instead of feature covariance to speed up off-policy prediction.
aiglean hits $300m arr as ai cost savings drive growth
enterprise ai search company glean reaches $300 million in annual recurring revenue, tripling in 15 months, with token cost reduction becoming a key selling point.
aifederated probe-logit distillation rate limits under heterogeneous bandwidth
new lower bounds show that the bandwidth term in federated probe-logit distillation is tight, and the method extends to nodes with different upload budgets.
aiai agent plans molecular tweaks for drug lead optimization
a new llm-based agent called trace uses tool planning to optimize drug-like molecules over multiple steps, improving properties while keeping key structures intact.
aibehavior-aware corrections stabilize off-policy td learning
replacing the auxiliary covariance matrix with the behavior bellman matrix improves stability in off-policy temporal-difference learning.
aianytime-valid federated conformal rag for llm swarms
a new method extends federated conformal rag to provide valid uncertainty estimates at any stopping time, enabling safer adaptive control in distributed language model systems.
ai