Student::printHeader(); for (int i = 0; i <= Roster::lastIndex; i++) { cout << classRosterArray\[i\]->getStudentID(); cout << '\\t'; cout << classRosterArray\[i\]->getFName(); cout << '\\t'; cout << classRosterArray\[i\]->getLName(); cout << '\\t'; cout << classRosterArray\[i\]->getEmail(); cout << '\\t'; cout << classRosterArray\[i\]->getAge(); cout << '\\t'; cout << classRosterArray\[i\]->getDays()\[0\]; cout << '\\t'; cout << classRosterArray\[i\]->getDays()\[1\]; cout << '\\t'; cout << classRosterArray\[i\]->getDays()\[2\]; cout << '\\t'; cout << degreeProgramStrings\[classRosterArray\[i\]->getDegree()\] << std::endl; 

} 

VOCs

• Integration of RWKV models in transformers (https://twitter.com/huggingface/status/1658054038879870977)
• MEGABYTE: Predicting Million-byte Sequences with Multiscale Transformers (https://arxiv.org/abs/2305.07185)
• Brex's Prompt Engineering Guide (https://github.com/brexhq/prompt-engineering)
• Uncensored Models (https://erichartford.com/uncensored-models)
• Open-LLMs - A list of LLMs for Commercial Use (https://eugeneyan.com/writing/open-llms/)
• The "Three-Hills" Model for evaluating GPT startups (https://assistedeverything.substack.com/p/the-three-hills-model-for-evaluating)
• Together’s $20M seed funding to build open-source AI and cloud platform (https://www.together.xyz/blog/seed-funding)
• A query language for programming (large) language models. (https://github.com/eth-sri/lmql)
• Report: OpenAI to Release Open Source AI Model (https://www.theinsaneapp.com/2023/05/openai-open-source-model.html)
• ChatHub is an all-in-one chatbot client (https://github.com/chathub-dev/chathub/blob/main/README.md)

• A guidance language for controlling LLMs (https://github.com/microsoft/guidance)
• OpenAI CEO Sam Altman tells senators he wants to see A.I. licensed. That might be good for us. It’s definitely good for OpenAI. (https://finance.yahoo.com/news/openai-ceo-sam-altman-tells-200241500.html)
• Advocating for Open Models in AI Oversight: Stability AI's Letter to the United States Senate (https://stability.ai/blog/stability-ai-letter-us-senate-ai-oversight)
• 3 Free Platforms for Personalized ChatGPT Experience (https://www.kdnuggets.com/2023/05/3-free-platform-personalized-chatgpt-experience.html)
• TinyStories: How Small Can Language Models Be and Still Speak Coherent English? (https://arxiv.org/abs/2305.07759)
• Symbol tuning improves in-context learning in language models (https://arxiv.org/abs/2305.08298)
• S-curves of LLMs (https://twitter.com/lessin/status/1658193420178030592)
• LLM University — Learning Resource for NLP (https://txt.cohere.com/llm-university/)

• Stability AI Releases StableStudio, the Open-Source Future of DreamStudio (https://stability.ai/blog/stablestudio-open-source-community-driven-future-dreamstudio-release)
• AI-powered coding, free of charge with Colab (https://blog.google/technology/developers/google-colab-ai-coding-features/)
• Summarize any Article as a Dialog Between two People with Bing Chat (https://www.aidemos.info/summarize-any-article-as-a-dialog-between-two-people-with-bing-chat/)
• Pandas AI — The Future of Data Analysis (https://medium.com/@fareedkhandev/pandas-ai-the-future-of-data-analysis-8f0be9b5ab6f)
• BratGPT: The evil older sibling of ChatGPT (https://bratgpt.com/)
• Hippocratic is building a large language model for healthcare (https://techcrunch.com/2023/05/16/hippocratic-is-building-a-large-language-model-for-healthcare/)
• Google’s newest A.I. model uses nearly five times more text data for training than its predecessor (https://www.cnbc.com/2023/05/16/googles-palm-2-uses-nearly-five-times-more-text-data-than-predecessor.html)
• Zoom will soon integrate Anthropic’s chatbot across its platform (https://www.theverge.com/2023/5/16/23725342/zoom-anthropic-ai-powered-chatbot-claude)
• Tom’s Hardware AI Chatbot (https://www.tomshardware.com/news/tomshardware-ai-chatbot)
• Numbers every LLM Developer should know (https://github.com/ray-project/llm-numbers)

• ChatGPT app for iOS (https://openai.com/blog/introducing-the-chatgpt-app-for-ios)
• CLI tools for working with ChatGPT and other LLMs (https://simonwillison.net/2023/May/18/cli-tools-for-llms/)
• All languages are NOT created (tokenized) equal (https://blog.yenniejun.com/p/all-languages-are-not-created-tokenized)
• Using LangChainJS and Cloudflare Workers together (https://blog.cloudflare.com/langchain-and-cloudflare/)
• BT to cut 55,000 jobs with up to a fifth replaced by AI (https://www.bbc.com/news/business-65631168)
• The AI takeover of Google Search starts now (https://www.theverge.com/2023/5/10/23717120/google-search-ai-results-generated-experience-io)
• Debt Collectors Want To Use AI Chatbots To Hustle People For Money (https://www.vice.com/en/article/bvjmm5/debt-collectors-want-to-use-ai-chatbots-to-hustle-people-for-money)
• MTIA v1: Meta’s first-generation AI inference accelerator (https://ai.facebook.com/blog/meta-training-inference-accelerator-AI-MTIA/)
• PaLM 2 Technical Report (https://arxiv.org/abs/2305.10403)
• Professor Flunks All His Students After ChatGPT Falsely Claims It Wrote Their Papers (https://www.rollingstone.com/culture/culture-features/texas-am-chatgpt-ai-professor-flunks-students-false-claims-1234736601/)
• Running chained LLMs with TypeScript in production (https://www.inngest.com/blog/running-chained-llms-typescript-in-production)
• Any model made available in the EU, without first passing extensive, and expensive, licensing, would subject companies to massive fines of the greater of €20,000,000 or 4% of worldwide revenue. Opensource developers, and hosting services such as GitHub... would be liable (https://twitter.com/jeremyphoward/status/1657920768771244032?cxt=HHwWgIDShfeJj4IuAAAA)

• Using ChatGPT for home automation (https://www.atomic14.com/2023/05/14/is-this-the-future-of-home-automation.html)
• Finetuning LLMs Efficiently with Adapters (https://magazine.sebastianraschka.com/p/finetuning-llms-with-adapters)
• Chatting with an Airbnb listing using data-backed generation (https://nokori.com/blog/2023-05-19-chatting-with-an-airbnb-listing)
• Google's AI Music Datasets: MusicCaps, AudioSet and MuLan (https://www.audiocipher.com/post/musiccaps-audioset-mulan)
• Intel Announces Aurora genAI, Generative AI Model With 1 Trillion Parameters (https://wccftech.com/intel-aurora-genai-chatgpt-competitor-generative-ai-model-with-1-trillion-parameters/)
• China is using AI to raise the dead, and give people one last chance to say goodbye (https://www.businessinsider.in/tech/news/china-is-using-ai-to-raise-the-dead-and-give-people-one-last-chance-to-say-goodbye/articleshow/100380496.cms)
• What I don’t like about chains of thoughts and why language is a bottleneck to efficient reasoning (https://samsja.github.io/blogs/cot/blog/)
• Girlfriend GPT is a Python project to build your own AI girlfriend using ChatGPT4.0 (https://github.com/EniasCailliau/GirlfriendGPT)
• Interact your data and environment using the local GPT (https://github.com/csunny/DB-GPT)
• He Helped Train ChatGPT. It Traumatized Him. (https://www.bigtechnology.com/p/he-helped-train-chatgpt-it-traumatized)
• Meta Introducing speech-to-text, text-to-speech, and more for 1,100+ languages (https://ai.facebook.com/blog/multilingual-model-speech-recognition/)
• LIMA: Less Is More for Alignment (https://arxiv.org/abs/2305.11206)
• airoboros: using large language models to fine-tune large language models (https://github.com/jondurbin/airoboros)
• DarkBERT: A Language Model for the Dark Side of the Internet (https://arxiv.org/abs/2305.08596)
• Bringing Open Large Language Models to Consumer Devices (https://mlc.ai/blog/2023/05/22/bringing-open-large-language-models-to-consumer-devices)
• LangChain AI Handbook(https://www.pinecone.io/learn/langchain/)
• Apple Restricts Employee Use of ChatGPT, Joining Other Companies Wary of Leaks (https://www.wsj.com/articles/apple-restricts-use-of-chatgpt-joining-other-companies-wary-of-leaks-d44d7d34)
• Welcome to the final phase of software eating the world. SaaS companies beware, you're on the menu this time (https://www.businessinsider.in/tech/news/welcome-to-the-final-phase-of-software-eating-the-world-saas-companies-beware-youre-on-the-menu-this-time-/articleshow/100418224.cms)

• Meta AI Unleashes Megabyte, a Revolutionary Scalable Model Architecture (https://www.artisana.ai/articles/meta-ai-unleashes-megabyte-a-revolutionary-scalable-model-architecture)
• RWKV: Reinventing RNNs for the Transformer Era (https://arxiv.org/abs/2305.13048)
• VanillaNet: the Power of Minimalism in Deep Learning (https://arxiv.org/abs/2305.12972)
• QLoRA: 4-bit finetuning of LLMs (https://twitter.com/Tim\_Dettmers/status/1661379354507476994)
• Baize v2 7B chatbot (https://twitter.com/XuCanwen/status/1661099080746143745)
• Against LLM maximalism (https://explosion.ai/blog/against-llm-maximalism)
• Microsoft Build brings AI tools to the forefront for developers (https://blogs.microsoft.com/blog/2023/05/23/microsoft-build-brings-ai-tools-to-the-forefront-for-developers/)
• Your A.I - Powered Quran Companion (https://www.tarteel.ai/)
• How GitHub Copilot is getting better at understanding your code (https://github.blog/2023-05-17-how-github-copilot-is-getting-better-at-understanding-your-code/)
• The White House wants your opinion on how to regulate AI (https://www.semafor.com/article/05/23/2023/white-house-ai-national-strategy)
• State of GPT by Andrej Karpathy (https://build.microsoft.com/en-US/sessions/db3f4859-cd30-4445-a0cd-553c3304f8e2)
• chatGPT plugin that decorates an empty room in different styles from a single photo (https://old.reddit.com/ChatGPT/comments/13qkr0j/i_made_a_chatgpt_plugin_that_decorates_an_empty/)

• Chatbot Arena Leaderboard Updates (Week 4) (https://lmsys.org/blog/2023-05-25-leaderboard/)
• Gorilla: Large Language Model Connected with Massive APIs (https://gorilla.cs.berkeley.edu/)
• AI Canon (https://a16z.com/2023/05/25/ai-canon/)
• PostgresML raises $4.7M to launch serverless AI application databases based on Postgres (https://postgresml.org/blog/postgresml-raises-4.7M-to-launch-serverless-ai-application-databases-based-on-postgres)
• The False Promise of Imitating Proprietary LLMs (https://arxiv.org/abs/2305.15717)
• Compress, Then Prompt: Improving Accuracy-Efficiency Trade-off of LLM Inference with Transferable Prompt (https://arxiv.org/pdf/2305.11186v1.pdf)
• GPTQLoRA: Efficient Finetuning of Quantized LLMs with GPTQ (https://github.com/qwopqwop200/gptqlora/tree/main)
• A GPT-Powered Chatbot For Any Site. (https://helphub.commandbar.com/)
• Create any UI with AI (https://dynaboard.com/blog/dynaboard-ai)
• Building Cloud Dev Assistants with GPT-4 and Pinecone on Gitpod (https://www.gitpod.io/blog/building-cloud-dev-assistants-with-gpt-4-on-gitpod)
• Incident Report for OpenAI (https://status.openai.com/incidents/jbt079x532bg)
• New superbug-killing antibiotic discovered using AI (https://www.bbc.com/news/health-65709834)
• How To Finetune GPT Like Large Language Models on a Custom Dataset (https://lightning.ai/pages/blog/how-to-finetune-gpt-like-large-language-models-on-a-custom-dataset/)
• PB reddit 7B - a lora trained on financial subreddits (https://huggingface.co/winddude/pb\_lora\_7b\_v0.1)
• Generative AI for Constructive Communication Evaluation and New Research Methods (https://ai4comm.media.mit.edu/)
• 58% of US adults familiar with ChatGPT, 14% have tried it (https://www.pewresearch.org/short-reads/2023/05/24/a-majority-of-americans-have-heard-of-chatgpt-but-few-have-tried-it-themselves/)
• OpenAI offers grants for developing democratic AI decision-making (https://openai.com/blog/democratic-inputs-to-ai)
• What is Project Tailwind and How it Works? How to Try Google Tailwind (https://openaitrend.com/what-is-project-tailwind/)
• Build and run useful autonomous agents (https://github.com/TransformerOptimus/SuperAGI)
• What is BLOOMChat and How to Use it? (https://openaitrend.com/what-is-bloomchat/)
• TaxGTP, the first AI tax assistant, introduced in UAE (https://fastcompanyme.com/news/taxgtp-the-first-ai-tax-assistant-introduced-in-uae/)
• It's not just statistics: GPT-4 does reason (https://jbconsulting.substack.com/p/its-not-just-statistics-gpt-4-does)
• Plug in and Play Implementation of Tree of Thoughts: Deliberate Problem Solving with Large Language Models that Elevates Model Reasoning by atleast 70% (https://github.com/kyegomez/tree-of-thoughts)
• The Age of Agents (https://sudoapps.substack.com/p/the-age-of-agents)

• ChatGPT conversations can be shared publicly (https://help.openai.com/en/articles/7943611-how-do-i-create-a-shared-link)
• GPT-4 in Minecraft (https://twitter.com/DrJimFan/status/1662115266933972993)
• What Neeva's quiet exit tells us about the future of AI startups (https://www.supervised.news/p/what-neevas-quiet-exit-tells-us-about)
• UAE's Technology Innovation Institute Launches Open-Source "Falcon 40B" Large Language Model for Research & Commercial Utilization** (https://www.tii.ae/news/uaes-technology-innovation-institute-launches-open-source-falcon-40b-large-language-model)
• All the Hard Stuff Nobody Talks About when Building Products with LLMs (https://www.honeycomb.io/blog/hard-stuff-nobody-talks-about-llm)
• IBM’s CEO expects A.I. to be so good at back office work that he plans to pause hiring humans for those jobs (https://fortune.com/2023/05/01/ibm-ceo-ai-artificial-intelligence-back-office-jobs-pause-hiring/)
• Leena AI Unveils WorkLM, it’s Proprietary Large Language Model (https://analyticsindiamag.com/leena-ai-unveils-worklm-its-proprietary-large-language-model/)
• Trends in AI (https://theaireport.substack.com/p/ai-report-3)
• Democratizing AI with open-source language models (https://lwn.net/Articles/931853/)
• Large Language Models are Few-Shot Health Learners (https://arxiv.org/abs/2305.15525)
• The Larger They Are, the Harder They Fail: Language Models do not Recognize Identifier Swaps in Python (https://arxiv.org/abs/2305.15507)
• Landmark Attention: Random-Access Infinite Context Length for Transformers (https://arxiv.org/abs/2305.16300)
• Lawyer cites fake cases invented by ChatGPT, judge is not amused (https://simonwillison.net/2023/May/27/lawyer-chatgpt/)
• Nvidia ACE Brings AI to Game Characters, Allows Lifelike Conversations (https://www.tomshardware.com/news/nvidia-ace-brings-npcs-to-life)
• ChatGPT seems to be trained on copyrighted books like Harry Potter (https://www.newscientist.com/article/2372140-chatgpt-seems-to-be-trained-on-copyrighted-books-like-harry-potte)
• AI tools like ChatGPT are built on mass copyright infringement (https://www.theglobeandmail.com/opinion/article-ai-programs-like-chatgpt-are-built-on-mass-copyright-infringement/)

• Prompt engineering techniques (https://learn.microsoft.com/en-us/azure/cognitive-services/openai/concepts/advanced-prompt-engineering?pivots=programming-language-chat-completions)
• NVIDIA Announces DGX GH200 AI Supercomputer (https://nvidianews.nvidia.com/news/nvidia-announces-dgx-gh200-ai-supercomputer)
• Seven new no-cost generative AI training courses to advance your cloud career (https://cloud.google.com/blog/topics/training-certifications/new-google-cloud-generative-ai-training-resources)
• Model Lab - A new tool to make sense of training LLMs (https://www.reddit.com/mlscaling/comments/13rvx0i/t\_introducing\_model\_lab\_a\_new\_tool\_to\_make\_sense/)
• AI Server Cost Analysis – Memory Is The Biggest Loser (https://www.semianalysis.com/p/ai-server-cost-analysis-memory-is)
• Winning the AI Products Arms Race (https://www.reforge.com/blog/ai-products-arms-race)
• How much would it have cost if GPT-4 had written your code? (https://pypi.org/project/cost-of-code/)

• Chain-of-Thought Hub: Measuring LLMs' Reasoning Performance (https://github.com/FranxYao/chain-of-thought-hub)
• AI Weirdness: the strange side of machine learning (https://www.aiweirdness.com/)
• We deserve to know if something was generated by AI (https://ntietz.com/blog/ai-text-should-be-labeled/)
• Making OpenAI Whisper faster (https://nikolas.blog/making-openai-whisper-faste)
• A Comprehensive Multilingual Dataset for Advancing Code Understanding and Generation (https://github.com/FSoft-AI4Code/TheVault)
• AI And Gaming In Cars! Nvidia And Jaguar Parternship (https://rizznews.substack.com/p/ai-and-gaming-in-cars-nvidia-and)
• Scaling Data-Constrained Language Models (https://arxiv.org/abs/2305.16264)

Child I, event #3

Nicholas Johnson KC, for the prosecution, continues to cross-examine Letby in the case of Child I. He moves on to the third incident, on October 14, 2015.
Mr Johnson says Letby does not refer to this incident in her statement. Letby, in her evidence, said she did not recall this night.
Letby rules out staffing levels, medical incompetence or staffing mistakes as a contributory factor in the collapse of Child I for this incident.

The staffing rota for October 13-14, 2015 is shown to the court, with Letby in room 1 as the designated nurse for Child I. Joanne Williams is the designated nurse for two other babies in room 1 that night.
Letby is asked to look at her nursing notes for that night. Mr Johnson says Child I was tolerating handling and 'tone appears improved', according to Letby's notes.
The notes add: 'At 0500 abdomen noted to be more distended and firmer in appearance with area of discolouration spreading on right-hand side, veins more prominent. Oxygen requirement began to increase, colour became pale...gradually requiring 100% oxygen...blood gases poor as charted. Clear air entry, slightly reduced on left, chest movement reduced...continued to decline. Reintubated at approx 0700 - initially responded well. Abdomen firm and sitended. Overall colour pale. Xrays carried out...resuscitation commenced as documented...night and day staff members present'
Letby says she cannot recall the discolouration now. She does not recall it moved, but it was spreading by getting larger.
NJ: "Where did you get the time of 5 o'clock from?"
LL: "I don't know. I don't know if it's from paper charts or memory."
Mr Johnson says if Letby had seen this, she would have escalated it to a doctor.
LL: "I can't comment on what time the doctor did come."

At 5am, Letby recorded that Child I's "abdomen was noted to be more distended" with an "area of discolouration" that was "spreading" and "veins more prominent".
Resuscitation on the infant was commenced at this time.
Mr Johnson asks Letby about this incident.
"I can't comment because I don't remember it now," she tells the court.
"Don't or won't?" Mr Johnson asks.
"Don't," says Letby.

Mr Johnson says almost 24 hours earlier, Child I was found "almost dead", and then this incident happened. He asks what Letby would have done.
LL: "I would have escalated it to someone, senior like a doctor."
Mr Johnson shows the doctor's note, which mentions: "Abdomen distended and mottled".
LL: "I can't say specifically what time I asked him to come, the note says he came at 5.55am".

Letby says "squeaky" air entry is not an emergency.
"I can't say now what time I called the doctors or who I escalated it to," Letby says.
"'Asked to see patient' is not very urgent," Mr Johnson says, referring to her notes.
Letby says there is no other way to call a doctor, other than an emergency call which she didn't think was warranted.

Mr Johnson says this would have been an emergency for Child I.
LL: "I don't believe it was an emergency, I believe it showed a decline."
NJ: "You sabotaged [Child I] at about 6 o'clock, didn't you?"
LL: "No."

A prescription chart shows Dr Matthew Neame prescribed morphine sulphate for Child I, and the infusion was commenced at 5.50am.
A fluid chart shows '0530 abdo distended++' in Letby's writing.
Letby says by 6am, the oxygen requirement had gone up to 100% for Child I, from 60% at 5am.
Letby had written 'squeaky' for the oxygen level at 5am. Letby tells the court this meant the air entry for Child I was not clear.
Letby says squeaky air entry is not an emergency.
Mr Johnson says there is also expanding discolouration and a distended abdomen.
Letby denies copying the word 'squeaky' for the 5am oxygen column from Dr Neame's 5.55am note. Letby: "I disagree."
Letby says she recalls Dr Neame saying the mottling was unusual; she cannot recall the mottling specifically.
A report showed Child I's gaseous distention of the bowel had increased on October 14 since the previous x-ray at October 13. Child I had been on a ventilator and nil by mouth.
Letby denies injecting air into Child I.
NJ: "You had inflated [Child I] with air, hadn't you?"
LL: "No."
At 7am, Child I had a significant desaturation.
Letby's note: 'Reintubated at approx 0700 - initially responded well. Abdomen firm and distended. Overall colour pale. Xrays carried out.'
NJ: "That is because you were sabotaging her, isn't it?"
LL: "No."
Letby says she does not remember the 7am desaturation "with any clear detail".

"Child I collapsed again at 7am, didn't she?" Nick Johnson, prosecuting, says.
"Yes," says Letby,
"This is one of those cases where air was going in and out of her but she was not oxygenating. Do you remember that?"
"Yes," says Letby.
The night shift handover commences at 7.30am, so at this point Child I was handed into the care of another nurse.
Letby continues to say she does not remember this shift.

Child I, event 4 (fatal collapse)

Mr Johnson moves to the final event for Child I, when she died on October 23.
Prior to that, Child I had been moved to Arrowe Park Hospital before returning to the Countess of Chester Hospital's neonatal unit.
Mr Johnson shows Letby observation charts for Child I from the previous day. Letby accepts Child I's observations were stable, save for one slightly raised respiration rate reading.
She agrees Child I was self-ventilating in air at this point. She accepts Child I's abdomen was, the previous day, soft and non-distended.
NJ: "Would you agree that despite three life-threatening events in the previous three weeks, [Child I] appeared to be in a stable condition?"
LL: "Yes."

For the night of October 22-23, Lucy Letby is a designated nurse for a baby in nursery room 2 and one in room 3. Ashleigh Hudson is the designated nurse for Child I in room 1 and one other baby.
Mr Johnson tells the court the baby in room 2 went to a hospital in Stoke during that night shift.
Letby says there were staffing issues, which were "not ideal", which were a contributory factor in the treatment of Child I following the collapse, in that a doctor had to be called away during the event.
Letby: "Considering what [Child I] had been through, she was a poorly baby, the doctors were not with her at all times...once she deteriorated."

"I think potentially there were staffing issues that may have affected her [the nurse on duty] but I don't know how," she says.
"The staffing was not ideal."
Mr Johnson asks if this is because she thought the colleague assigned to Child I was "not up to the job".
Letby says no, and says she means because "the doctor had to be called away to another delivery" in the labour suite.
"You didn't like being in nursery three though," Mr Johnson asks.
"No, I disagree," says Letby.
He then asks if she was jealous of her colleague for being in a higher dependency nursery.
"No, I have nothing to be jealous of," says Letby.

Letby adds she believed Ashleigh Hudson was capable of looking after Child I, for Child I's nursing needs at this stage.

The neonatal schedule for that night is shown to the court.
Letby sent a message on October 22 at 8.47pm to a colleague: '...Unit nice. Transport on way to take my baby back to Stoke. Only 8 babies. Off duty not out. X'
Mr Johnson says this refers to the baby he mentioned earlier who was transferred out during that night.
The court hears that transfer process, noted as completing at 1am, is not a 'five-minute' process, and takes time and involves family communication.

Child I collapsed at 11.57pm.
Letby denies falsifying a note for the Stoke-transfer baby prior to that at 11pm.
The court is shown a nursing note by Ashleigh Hudson, which the court heard was timed at 10.57pm. 'Longline removed due to constant occlusions; neonatal nurse Lucy Letby unable to flush, so Paeds Reg Rachel Chang informed.'
Dr Chang had written, for the Stoke-transfer baby at 10pm, the baby was safe for transfer.
Letby's note for this baby was written at 10.50pm, and completed at 10.52pm. It included a documentation of a longline infusion with a 10% dextrose fluid. Letby has co-signed the document.
Mr Johnson says the 'original 2300' reading has been changed to '2400' by Letby.
Letby said the '2300' reading was an error and it was changed to '2400' as the correct time. She adds: "The charts are there for everybody to look at."

This transfer, Letby says, would not have been a five-minute job and would have required a handover with the transport team and speaking to the family.
"I am going to suggest you were dealing with [the other baby] at 11pm," Mr Johnson says.
"Do you remember falsifying a recording relating to [the other baby]?"
Letby refutes this.
The court is then shown records relating to the baby in Letby's care. A close-up image of a medical note shows this baby being given a dextrose infusion at 00.00 (midnight), prior to the transfer.
But, the prosecution says, the four has been changed from a three, and the actual time this took place was 23.00 (11pm). They are accusing Letby of changing the notes to give herself an alibi in the lead-up to Child I's collapse.
"No I would not have changed a record, that was obviously written in error," Letby says. She says the note would have been signed off by her colleague.
The prosecution says she could have changed the time after it was signed.

Letby denies falsifying a fluid balance chart for the Stoke-transfer baby.
Mr Johnson asks if Letby recalls what Ashleigh Hudson said for the 11.57pm desaturation. He says Ms Hudson gave evidence to say Child I was crying, making a noise she had not heard before, different to a cry for hunger.
Letby: "I did not hear that cry at that point. When I entered that nursery, she was quiet...and apnoeic."
Letby says for this event, it was a case where one of the three nurses on duty that night would have had to come and assist in room 1.
Letby says she does not recall Ashleigh Hudson going to call for Child I's parents.

Nick Johnson, the prosecution barrister, asks Letby if she remembers saying she had "no recollection" of Child I, other than what appeared in the notes.
"I don't remember this baby without the notes," Letby agrees.
Letby's colleague said before Child I collapsed she had "been very unsettled" and made a sound she had never heard before.
Her colleague described this as a "loud, relentless,s almost constant with no fluctuation, cry" that was "very different to a hunger cry".
"This is another case of you gravitating to nursery one when you were in other less acute nurseries," Mr Johnson asks.
Letby refutes this was deliberate and says one of the three members of staff "would have had to assist".
She was involved in giving Child I medication - something she says she cannot remember but is backed up by the nursing notes.
The court is then shown more nursing notes, where the times have clearly been changed - with digits overwritten.

Letby says there is an error on the IV chart, and the time has changed.
Mr Johnson: "Three different mistakes on two different babies?"
Letby says she does not know who wrote in the different times.
NJ: "How do those sorts of mistakes happen?
Letby says when the unit gets busy, "we" can make errors on the paperwork.
NJ: "We? Or you?"
LL: "I don't believe it would have been me - we would both have been there for it."
NJ: "Or is it you altering medical records to put some time between you and serious events for [Child I]?"
LL: "No."
Letby adds: "I did not deliberately falsify any paperwork."

At 1.06am, Child I was crying again, the court hears.
Letby recalls Child I was crying, but cannot recall being there by the cotside first. She accepts she was in the nursery.
Mr Johnson asks if Ashleigh Hudson was called over by Letby.
LL: "She might have been in the nursery when I called her, I couldn't say."
Letby adds she could have "come in" [as her defence statement says] from the other part of the nursery.
Letby says she had her hands in the incubator, "trying to settle [Child I]."
LL: "My assessment of [Child I] at that time was she was hungry and rooting."
NJ: "You had pumped her full of air?
LL: "No."
NJ: "You were doing your best to kill her?"
LL: "No."
Letby: "I have never injected air into any baby."

NJ: "Do you remember interrupting [Child I]'s mother?"
LL: "No."
Mr Johnson says Child I's mother, in agreed evidence, recalled Letby was "smiling" and had talked about how Child I had been going on about 'enjoying' her bath. Child I had been bathed as part of the bereavement process following her passing.
NJ: "Why did you say that?"
LL: "It's trying to, in that awful situation - it wasn't meant with any malice. We still talk to them and treat them as if they were alive. It wasn't joking or...malice, it was trying to reflect on a happier memory."
NJ: "How can you say such things?"
LL: "She had her first bath when she was alive and that was what she had enjoyed, not the one when she had passed away."
NJ: "How do you know it was her first bath?"
LL: "Because I was there, we took photographs, it was a big occasion."
Mr Johnson asks how many baths Child I had in Arrowe Park - Letby says she cannot say.
NJ: "You were getting a thrill out of the grief and despair in that room, weren't you?
LL: "Absolutely not."

Child J

Copies of Letby's defence statement, edited to the relevant parts in the case by agreement, are handed out to members of the jury.
Mr Johnson moves to the case of Child J, a baby girl born on October 31, 2015.
Letby, in her defence statement, said she had never seen a baby with stomas before at the Countess, and other doctors were "equally unsure about stomas", and the parents were "more proficient" than the Countess staff at dealing with stomas.
A handover sheet was taken to Letby's home, the defence statement adds, unintentionally.
The night shift rota for November 26-27 is put up. Two band 4 unit nurses are named in the rota.
Child J was in room 4. The designated nurse was Nicola Dennison, a band 4 nurse also looking after one other baby in room 4.
Letby was the designated nurse for two babies in room 3 that night.
Letby rules out staffing levels as a contributory factor in Child J's collapse, nor medical incompetence, nor staffing mistakes.
Letby accepts the evidence from Child J's mother that Child J was well and "about to go home in a day or two".

Letby adds there was an issue with Child J's stoma care, as it had been discussed among the nurses that they had little experience. She adds she does not "want to name names" on any specific nurses' lack of experience.
Mr Johnson refers to Nicola Dennison's previous experience with stomas, which she said in evidence she had experience of it.

Letby says over the years, she did not recall any other babies with stomas.

Mr Johnson says band 4 nurses, as said by Letby in evidence on May 16, should not be involved in stoma bag care, as they would be unfamiliar with the procedure. Letby said: "The unit was very busy and we had to use staff where we could."
Letby says she was not referring to Nicola Dennison specifically, but the nursing situation overall.
She says there was not an issue over staffing levels at the time of Child J's collapse.
Asked to explain a text message she had sent to a colleague Letby tells the court: "Sometimes I felt nurses would take on roles which I didn't think they were trained enough in".
The next message adds: "It's shocking really that they are willing to take the responsibility for things that they have no training or experience etc on. Don't think they appreciate the potential difficulties X"
Letby agrees she believed it was "potentially dangerous".
Mr Johnson says the impression of the court was that band 4 nurses were not qualified in stoma care, and the hospital was 'cutting conrers' by assigning such nurses to those tasks. Letby agrees.
The court is shown a document about the duties for special care babies [such as Child J], which includes stoma care.
LL: "You need to appreciate the context that the unit was not familiar with stomas."
NJ: "This nurse was familiar with stomas, wasn't she?"
LL: "In her opinion, yes."
Mr Johnson says Letby was deliberately creating the impression to the jury that the care for Child J was deficient.
LL: "I do think that. I don't think she had a high standard of care.
"I don't think anyone was overly confident in saying 'I know what to do with a stoma'. We were led by the parents..."
Mr Johnson asks why Nicola Dennison was not challenged about this.
LL: "I can't answer that."

Letby denies not being happy in nursery room 3, or being happier in nursery room 1.
Letby accepts the evidence of Nicola Dennison that babies in room 4 should have the light off overnight, as they are due to go home.
Mr Johnson says two pairs of events for child J happened; one pair in room 4, one pair in room 2.
The room 4 incidents happened at 3am and 4.57am, and the room 2 incidents happened either side of 7am. Letby accepts this was the case.
The court is shown a night shift staffing rota at the end of the night, in which Child J was in room 2.
Letby is asked if she has any memory of the earlier pair of incidents. She says she does not have a recollection.
She says from her memory, Child J had a seizure and was moved to room 2. She says she could be mistaken in her memory.

The court is shown an apnoea/brady/fit chart for Child J on November 27, recording events for Child J at 4.40am and 5.03am, recorded by nurse Nicola Dennison, in nursery 4.
The desaturations are recorded by Dr Kaliyilil Verghese.
Letby recalled when she was called in to room 4, Child J was 'fitting', not desaturating.
Letby accepts that by 6.28am, Child J had been moved to room 2, as a text message written by her to a colleague had said that was the case.

The court is then shown messages Letby sent to her colleague early in the morning.
Colleague [6.24am]: Good night? X
Letby [6.28am]: No [Child J] in 2 screened had profound de sats and just got 32-week twins born at home one with cleft lip and only 5 staff! X
The prosecution alleges this was evidence Letby was in nursery two with Child J, while her colleagues concentrated on the twins.
"I can't recall where I was when I sent that message," Letby says.

Letby says she cannot recall where she was when she sent the 6.28am message, whether she was in room 2 or not.
The message added: 'only 5 staff!'
Mr Johnson: "So it was all hands to the pump then?" as twins had been admitted to room 1 as an emergency.
LL: "Yes."
Mr Johnson says all staff would have been concentrated in room 1.
LL "Not all, but most, yes."
NJ: "You were not involved in that, were you?"
LL: "Not from memory, no."
Mr Johnson says the message sent at 6.31am would have meant Letby would not have been in room 1. Letby agrees.
Letby says Mary Griffiths would not have been in room 1 as she was not an intensive treatment unit-trained nurse. She denies she would have been the last nurse for room 2.
Letby accepts, from looking at the neonatal schedule, she would have been in room 2 when the emergency twins were admitted to room 1.
NJ: "There would have been a lot of distractions...wouldn't there?"
LL: "I don't know what you're implying."
NJ: "The medical staff would have their attention focused on the twins, and any help that could be spared would have gone on the twins.
NJ: "Do you accept that a lot of help was needed?"
LL: "It would be normal practice to get in the consultant when we only had the registrar, yes."
Dr John Gibbs arrives at 6.34am, earlier than normal for his shift, to assist.

The last message Letby sent to her colleague was 6.49am. The colleague sent three messages which were not replied to in the following minutes.
NJ: "That's because you were in nursery room 2, sabotaging [Child J], weren't you?"
LL: "No, I wasn't."
Letby accepts that, on the neonatal schedule, she is not recorded doing anything in the half hour prior to Child J's collapse at 6.56am.
Mr Johnson refers to Dr John Gibbs's notes of 'sudden desats (to unrecordable levels) at 6.56 and at 7.24 and bradycardia. Both associated with clenching of hands, stiff limbs, and on second occasion, eyes deviated to left.'
NJ: "This was your doing?"
LL: "No, it wasn't."
Letby accepts it was an emergency situation and Dr Gibbs had to be called away from room 1 to Child J in room 2.
NJ: "You took your opportunity, when all the resources at the NNU were concentrated on the twins who had been admitted as an emergency."
LL: "No."
Letby accepts evidence had been heard saying there was no known cause for Child J's deterioration.
Letby had care of Child J the following night, which the court is shown, from Letby's notes for that night, 'nothing happened'.

The following night, nothing happened to Child J - while Letby has tried to claim this is evidence she didn't hurt her (because why would she hurt her on one shift and not another), the prosecution points out that the following night Child J's parents were present on the unit.

Child K

Mr Johnson moves on to the case of Child K, born on February 17, 2016.
Letby said, in her defence statement, she did not recall the events of February 17, and did not recall saying to Dr Ravi Jayaram that Child K had just started deteriorating.
She said she had done nothing to interfere with Child K's tube or the alarm.
She added the Countess neonatal unit was not capable, given its staffing levels, of looking after a baby of Child K's gestational age.
Letby tells the court she has no memory of such a conversation with Dr Jayaram. She says it is "difficult" to dispute Dr Jayaram's recollection of the event as she had no memory of it.
She denies she has changed her version of events since starting to give evidence.

Letby is asked if she understands the reason why Child K was born at the Countess.
LL: "Yes."
Mr Johnson tells the court it was deemed 'too risky' to transfer Child K and her mother to another hospital at that stage, and that was why Child K was born at the Countess.
LL: "I don't know why more effort was not made to find a bed for her [elsewhere]."
NJ: "You have persistently given the impression that the Countess has taken on babies it [is not able to look after and that is why they collapse]."
LL: "Yes."
NJ: "Is that the reason you said to the jury you didn't understand why [Child K] was born at the Countess?"
LL: "I don't understand why she was born at the Countess."
NJ: "Is it to bolster your defence?"
LL: "No."
LL: "I understand why she was born there but I don't necessarily agree [with the decision to have her born there]."

Letby says she does not recall the latter two desaturations for Child K, and does not accept Dr Jayaram's evidence in the first desaturation.
Mr Johnson says he will deal with these in a different order than chronologically; he will cross-examine on the second desaturation first.
Letby says she does not know what happened to Child K, so does not cite staffing levels as a contributory factor in Child K's desaturations.
She says she feels "potentially" the ET Tubes were not secured for Child K.

The second desaturation occurred at 6.10-6.15am on February 17, 2016.
The court hears a note on Child K's birth and assessment was typed up by Letby on a computer from 6.04am-6.10am. The note would have been taken from paper charts taken by the cotside.
NJ: "You were at [Child K's] cotside a minute or two before she desaturated, didn't you?"
Letby says she would have got the notes from the cotside "at some point" prior to her typing them up.

Mr Johnson then asks about the admission form Letby filled out. She says she got these from the cot side and took them to the computer.
"I think I know where you are going, we will dance the dance if you want to," Mr Johnson says.
Letby's defense barrister objects to this, saying comments like this are "belittling". The judge agrees and asks Mr Johnson to refrain from such remarks in future.
Letby is accused of moving Child K's tube, causing her to collapse.
"You moved Child K's tube when you took those notes back to her cot side," Mr Johnson asks.
"No I did not," says Letby.
Court is now adjourned for the day. The trial will resume at 10.30am on Monday

Detroit Tigers

Toledo Mud Hens

Erie SeaWolves

West Michigan Whitecaps

Lakeland Flying Tigers