reasoning · 77 topics

ox alpha

ox alpha

empero ai qwythos 9b claude mythos 5 1m gguf

exploring self distilled reasoning for supervised fine tuning with amazon nova

glm 5 2 open weights model

claude fable game generation

claude fable 5

glm 5 1 reasoning 1m cleaned

claude opus 4 6 dataset 10k

mathnet multilingual olympiad dataset

openai gsm8k

opus 4 6 reasoning 3000x filtered

ornith 1 5 beats 35b models on reasoning

kyutai voice of reason stitch 9b

morethought fable 51 max reasoning filtered 5000x

ant group ling 3 0 flash vl

openai gpt 6 astra cracks epoch s hardest math benchmark in 14 months

meta muse spark 1 3

nvidia nemotron ioi 2026

faunix qwen38 27b distillation 40k

gemini 3 8 flash and 3 8 flash cyber

multiverse computing quasar 438b

sapiens ai agnes 2 5 pro beta

epochs ebr bench reveals humans learn while frontier ai stays stuck

z ai ox alpha model

granite 4 2 llms

qwen3 8 max glm5 2 kimi k3 distillation

glm 5 3 artificial analysis benchmarks

qwen3 8 27b scores 52 on artificial analysis

models are getting dumber on purpose

upstages solar pro 4 jumps 28 points to beat human agents

deepseek v4 flash 0731

arbitrage efficient reasoning

openai gpt 5 6 sol

inclusionai ling 3 0 flash

openai astra math

qyrou reasoning corpus 4k 5m v1

is ai reasoning right for the wrong reasons

qwen38 max distillation 50k

claude opus 5

lead breaking the no recovery bottleneck

gpt 5 6 claude fable 5 muse spark 1 1

controlling reasoning effort in llms

schema harness arc agi 3

gpt 5 6 family

making claude a chemist

deepmind erdos problems solved

openai model disproves discrete geometry conjecture

qwen 3 7 dropped on qwen chat

hermes agent reasoning traces

modotte codex 2m thinking

google deepmind ai co mathematician frontiermath

teaching claude why

natural language autoencoders

proteo r1 reasoning protein design

nvidia nemotron 3 nano omni

harvard o1 er diagnosis

nvidia nemotron 3 nano omni 30b a3b reasoning bf16

grok 4 3 benchmark performance

jackrong deepseek v4 distill 8000x

2026 04 27 papers 2604 22074

stepaudio 3 realtime technical report

negative self distillation reasoning

building multilingual bridges data mixing as the pillar of generalization for in language reasoning

transformer logical consistency qk alignment

flowbalance verifier grounded self improvement

random attention rethinking kv cache eviction for efficient reasoning

multimodal reasoning sycophancy measurement

evolution strategies for llm reasoning broader coverage

criticl weak to strong generalization

trace temporal reasoning benchmark

ai4ai test time capability transfer

bdh cq in context learning with recurrent latent reasoning

llm reasoning reliability reproduction

slpo scaling latent reasoning

evothink evolving thinking in large reasoning models

2507 12549 serial scaling