Amazon Science ✓

@AmazonScience · Global · joined 10 Dec 2019

The latest news and research from Amazon's science community. #AmazonScience

92 103Followers

1 917Following

7 626Posts total

97KViews on collected posts

Posts recentes

Amazon Science ✓@AmazonScience4 Sep

.@Amazon and @DARPA brought together 150+ researchers in Seattle last week, with speakers including @awscloud CEO @mattsgarman, Fields Medalist @TaoistTerence, @EPrinceton Associate Professor @BorisHanin, and @NSAGov's Michael O'Hara to explore how AI is transforming mathematical

10.8K views · 22 likes · 3 reposts · 2 replies Open on X →

Amazon Science ✓@AmazonScience2 Sep

Amazon Redshift researchers were awarded Best Paper Runner-Up: Industrial Track at @VLDBconf for eliminating compilation cold starts in query execution – cutting compilation time from seconds to milliseconds with a 7x speedup on TPC-DS benchmarks. #VLDB2026 2.4K views · 12 likes · 0 reposts · 0 replies Open on X →

Amazon Science ✓@AmazonScience1 Sep

@AmazonScience @blackenedgold @InferenceLang provides native compilation to WebAssembly with auto formalization to @CoqLang out of the box. Currently, proof holes auto completion is in closed beta but will be available soon 276 views · 1 likes · 0 reposts · 0 replies Open on X →

Amazon Science ✓@AmazonScience31 Aug

@AmazonScience spec is still a trap 2.7K views · 4 likes · 0 reposts · 0 replies Open on X →

Amazon Science ✓@AmazonScience31 Aug

@AmazonScience the proof only says the code matches the spec, so the whole thing rests on the spec being what you actually wanted. it pays off where the spec is way shorter than the code, which is what isolation primitives are. gets a lot worse the further up you go 2.2K views · 5 likes · 0 reposts · 0 replies Open on X →

Amazon Science ✓@AmazonScience31 Aug

Verus is an open-source, automated program verifier for Rust that mechanically checks code against a formal mathematical specification for all possible inputs. Amazon used it to prove correctness of Nitro Isolation Engine primitives. https://t.co/SuFeU5g8nf 73.9K views · 541 likes · 70 reposts · 8 replies Open on X →

Amazon Science ✓@AmazonScience28 Aug

When LLM judges agree, the right question is why. Shared prompts, model families, or training lineage can make a majority look stronger than it is. Dependence-aware aggregation via Ising models accounts for this, improving accuracy 9–14% over weighted majority vote. 2.9K views · 14 likes · 5 reposts · 0 replies Open on X →

Amazon Science ✓@AmazonScience25 Aug

How did a model upgrade make agents worse? By pairing real enterprise SOPs with functioning tools and ground-truth grading across 12 industries and 2,000+ tasks, SOP-Bench helps find such anomalies. https://t.co/B2NarMII2B 1.7K views · 2 likes · 0 reposts · 0 replies Open on X →

Em comparação com contas do mesmo porte

8 posts dos últimos 90 dias, ao lado da faixa de 10K–100K seguidores. alcance comum para o seu porte, reação mais fraca que a maioria.

Mediana de visualizações2 582esta conta924mediana para 10K–100K

Alcance, %2.80%esta conta3.62%mediana para 10K–100K

Engajamento, %0.32%esta conta1.52%mediana para 10K–100K

Métrica	Esta conta	Mediana para 10K–100K	Proporção
Mediana de visualizações por post	2 582	924	2.79×
Alcance (visualizações ÷ seguidores)	2.80%	3.62%	0.77×
Taxa de engajamento	0.32%	1.52%	0.21×

Outras contas desta faixa → Comparar com outra conta → Como estas referências são construídas →

Growth & engagement

How the posts we collected actually performed: views and reaction rate post by post, what the audience did with them, and where the follower count goes.