LIVE PULSE
5.0 Anthropic CEO Amodei calls for slower AI development and shared safety rules — 11 src2.7 Agility Robotics unveils Digit 5 humanoid for warehouses and factories — 2 src2.5 Apple ships rebuilt Siri with Google Gemini, but not in the EU — 2 src2.2 Siri AI in macOS 27 Golden Gate: FAQ, Germany availability, privacy questions — 2 src1.8 Sam Altman says OpenAI will not go public in 2026, citing AI safety concerns — 5 src1.4 OpenAI contractors review real ChatGPT conversations to rate responses, report says — 2 src1.4 Anthropic data retention policy prompts firms to limit Claude use for sensitive work — 1 src1.4 VoiceCodeBench arXiv paper proposes benchmark for exact structured-token recovery in speech recognition — 1 src1.4 Arabic-Russian Parallel Corpus and LLM Benchmark for Scientific Text — 1 src1.4 Study Analyzes Self-Reported Limitations in NLP Research — 1 src5.0 Anthropic CEO Amodei calls for slower AI development and shared safety rules — 11 src2.7 Agility Robotics unveils Digit 5 humanoid for warehouses and factories — 2 src2.5 Apple ships rebuilt Siri with Google Gemini, but not in the EU — 2 src2.2 Siri AI in macOS 27 Golden Gate: FAQ, Germany availability, privacy questions — 2 src1.8 Sam Altman says OpenAI will not go public in 2026, citing AI safety concerns — 5 src1.4 OpenAI contractors review real ChatGPT conversations to rate responses, report says — 2 src1.4 Anthropic data retention policy prompts firms to limit Claude use for sensitive work — 1 src1.4 VoiceCodeBench arXiv paper proposes benchmark for exact structured-token recovery in speech recognition — 1 src1.4 Arabic-Russian Parallel Corpus and LLM Benchmark for Scientific Text — 1 src1.4 Study Analyzes Self-Reported Limitations in NLP Research — 1 src
HEATPULSEAI MAGAZINES
FLIP · FOLLOW · SAVE
Log inSign up

#post-training

7 curated events
papersTODAY 04:00 UTC

arXiv Paper Examines How Much Training Data Matters in On-policy Distillation

A new arXiv preprint investigates how much of the benefit from on-policy distillation actually comes from the training data used. Testing the two teacher-student pairings most often seen in practice, the authors report findings that challenge assumptions about data's role in the method. The work is positioned as a closer look at a technique that has become standard in frontier post-training pipelines.

distillation/post-training/llm/research
papersTODAY 04:00 UTC

ReWeight Uses Human Demonstrations and Sample Weighting for VLA Post-Training

A new arXiv preprint proposes ReWeight, a technique for post-training vision-language-action models when in-domain robot demonstrations are scarce. The approach retrieves relevant egocentric human demonstrations and applies sample weighting to make better use of that human data, since collecting robot-specific data is expensive. The paper targets adapting VLA models to particular robots and tasks.

vla/robotics/human-data/post-training
papersSEP 11 04:00 UTC

arXiv paper proposes statistical guarantees for post-training hyperparameter selection

A new arXiv preprint addresses how to choose hyperparameters after a model has already been trained, such as inference-time settings and implementation-level options. The work aims to move this process beyond ad hoc tuning by providing statistical validity guarantees for the selected configuration. It targets deployment scenarios where pre-trained models still have multiple degrees of freedom that need to be fixed.

research/hyperparameters/post-training/statistics
papersSEP 10 04:00 UTC

Paper proposes averaging SFT and RL policy logits for LLM post-training

An arXiv paper introduces a post-training method that blends supervised fine-tuning with reinforcement learning by averaging the output logits of a frozen reference policy with those of the trainable policy. The technique is integrated into GRPO, letting a model retain its supervised behaviors while still improving on verifiable rewards during reinforcement learning.

llm-training/reinforcement-learning/grpo/post-training
papersSEP 10 04:00 UTC

Direct Diversity Optimization for Diverse Successful Trajectories in Preference Post-Training

A new arXiv study addresses a limitation of post-training LLM agents with trajectory-level outcome labels: such supervision offers little signal for keeping multiple distinct successful strategies that branch from the same decision state. The authors frame this as a successful trajectory diversity problem and introduce Direct Diversity Optimization, a method intended to preserve varied winning paths during preference-based post-training.

agents/post-training/diversity/llm
papersSEP 12 04:00 UTC

LOCUS: Task-Aware Low-Rank Post-Training for Token-Efficient Language Generation

A new arXiv paper examines how the way post-training updates are parameterized influences the length of text that large language models produce. The authors propose LOCUS, a task-aware low-rank adaptation method intended to curb the verbosity that standard preference alignment tends to introduce without sacrificing usefulness. The work targets serving costs, which grow with output length.

arxiv/llm-efficiency/post-training/low-rank
papersSEP 12 04:00 UTC

Open recipe targets IMO gold with post-trained Nemotron math models

A new arXiv paper examines how post-training choices and test-time inference setups influence a model's ability to write natural-language proofs for difficult olympiad problems. Using Nemotron 3 Ultra as a base, the authors produce two specialist checkpoints via supervised fine-tuning and reinforcement learning, and release the training approach publicly.

math/post-training/open-source/benchmarks