Skip to main content

Improve your summarising - ChatGPT prompt


One of the most popular uses of Generative AI is summarising documents and information sources. Now for the but, could it be even better?

Adams et (2023) the issues of using ChatGPT4 to produce the right amount of information in a summary. Balancing the detail in the summary without it becoming too dense, it is hard to follow, whilst ensuring it is relevant to the source. In the paper they investigated applying a technique they called "Chain of Density" (COD) to produce a better summary compared to a standard "summarise is this" prompt.  

Chain of Density in Summary
Start with the standard summarise this kind of prompt - fairly non-specific.
1. Identify one to three more informative entities from the article
2. Write a new denser summary of the same length including all the entities identified (don't remove entities when identified) expanding the detail

Repeat steps 1 and 2 around 5 times. Applying guidelines around improving on the previous summary but must be self-contained (understood without the article),



To try this out I applied it to a paper
The prompt (below in bold) used was very slightly modified version from (Adams et al, 2023)
You will generate increasingly concise entity-dense summarises of the article (Article)loaded in. Repeat the following 2 steps 5 times Step 1. Identity 1-3 informative Entities (“;” delimited)from the article which are missing from the previous generated summary. Step 2. Write a new denser summary of identical length which covers every entity and detail from the previous summary plus the Missing Entities. A Missing Entity is:- -Relevant: to the main article. -Specific: descriptive yet concise (5 words or fewer). -Faithful: present in the Article. -Anywhere: located anywhere in the Article. Guidelines: -The first summary should be long (8-10 sentences, ~250 words) yet highly non-specific, containing little information beyond the entities marked as missing. Use overly verbose language and fillers (e,g. “this articles discusses”) to reach ~250 words. -Make every word count: rewrite the previous summary to improve flow and make space for additional entities. -Make space with fusion, compression, and removal of uninformative phrases like “the article discusses”. -The summaries should become highly dense and yet self-contained, e.g. easily understood without the Article. -Missing entities can appear anywhere in the new summary. -Never drop entities from the previous summary. If space cannot be made, add fewer new entities. Remember , use exact same number of words for each summary.


Produces 
Version 1

In an era where multimedia applications, especially video streaming services, are burgeoning due to the swift advancement and proliferation of user devices and internet services, ensuring quality of service (QoS) and quality of experience (QoE) for end-users presents a formidable challenge for service providers and network administrators. The document delves into the complexities surrounding multimedia traffic transmission over networks, highlighting the issues of bandwidth, delay, jitter, and loss ratio that often impede the delivery of high-quality video streaming services. It proposes an innovative solution that marries the principles of reinforcement learning with software-defined networking (SDN) to create a dynamic, intelligent routing framework capable of navigating these challenges. The framework aims to optimize routes for multimedia traffic, particularly video streaming, by learning from network conditions and making informed routing decisions to enhance end-user satisfaction. Simulation results underscore the effectiveness of this approach, demonstrating improvements in viewing quality, throughput, and stability of video quality during streaming, compared to existing techniques. The involvement of various academic institutions and researchers underscores the collaborative effort to address these network challenges, with the ultimate goal of refining multimedia streaming experiences in an ever-evolving digital landscape.


Fifth version

Leveraging the power of SDN and reinforcement learning, this paper advances multimedia traffic routing, focusing sharply on improving video streaming under diverse traffic conditions. It strategically employs machine learning for insightful analytics in network optimization, guiding real-time decisions to mitigate common QoS challenges. This intelligent routing framework ensures robust network performance, enhancing video delivery mechanisms to provide consistent, quality streams. By dynamically adapting routes to optimize for bandwidth and latency, it sets a benchmark in multimedia service provision, prioritizing user QoE. Simulation evidence underscores its efficiency, highlighting its superiority in ensuring high-quality video streaming experiences against traditional routing methods. This approach not only demonstrates the practical application of analytics and machine learning in network management but also marks a significant stride towards achieving unparalleled consistency and quality in multimedia streaming services, even amidst fluctuating network traffic.



Comment.
Both are fine, but personally I believe the final is a more informative read. But have a go yourself cut and paste the prompt in bold and put an article through it. If you are using the free version of ChatGPT or Gemini you might change the first line to say that the Article is going to be entered in the prompt that follows and then enter the text in the following prompt. Have fun and please share your views in the comments.



Reference

From Sparse to Dense: GPT-4 Summarization with Chain of Density Prompting

arXiv:2309.04269v1 (Preprint)


All opinions in this blog are the Author's and should not in any way be seen as reflecting the views of any organisation the Author has any association with. Twitter @scottturneruonn
Labels:

Comments

Post a Comment

Popular posts from this blog

Using an LLM to Find Themes using Thematic Analysis in an Academic Career

B rief recap In an earlier post, “Same prompt, four AIs: why answers aren’t always the same” , I looked at what happened when the same prompt was given to four different LLMs. Unsurprisingly, perhaps, the answers were not identical. That raised an interesting follow-on question: what was the prompt actually trying to do? The answer is that it was trying to support a form of thematic analysis. In that case, the object of analysis was an academic profile, including my Google Scholar profile. The version I discuss here is a modified prompt, used with Claude.ai, where I uploaded a CV and asked the system to find other appropriate public resources connected with a named person and institution. The aim was not simply to summarise the CV, but to identify visible and less visible themes across a whole career. Why use thematic analysis? Thematic analysis is widely used by academics, especially in qualitative research. Braun and Clarke’s well-known paper, Using thematic analysis in psy...
Post a Comment
Read more

Prompt Engineering, Context Engineering and Agentic AI in Higher Education: A Short Briefing

AI in Education Prompt engineering, context engineering and agentic AI are often used interchangeably, but the literature treats them as distinct. Prompt engineering concerns crafting effective single instructions to a model (Glean, 2026). Context engineering is the broader discipline of designing and managing the entire informational environment around a model—memory, retrieval, tool outputs and conversation state—rather than a one-off instruction (Abstracta, 2026); prompt engineering operates within the context window, while context engineering determines what fills it (arXiv:2606.12422, 2026). Agentic AI describes systems that plan and execute multi-step tasks with delegated autonomy, raising organisational questions of accountability rather than purely technical ones (MIT Sloan, 2026; Palo Alto Networks, 2026). There is genuine debate about prompt engineering's durability. IEEE Spectrum (2025) reported research suggesting prompting is increasingly performed by models themse...
Post a Comment
Read more

AI as a Mirror: Transforming Vague Student Ideas into a More Rigorous Project Agreement

The Problem: The "Generic App" and the "Time Sink" We’ve all been there: a student walks into a 1-to-1 with a vague desire to "do something with AI" or "build a fitness app." You spend 45 minutes trying to find a technical "hook" that justifies a Level 6 or Level 7 grade, only for the student to drift back into "CRUD app" territory by week three. The Philosophy: AI as a Mirror Instead of you doing the heavy lifting, this workflow uses AI as a Mirror . It reflects the student’s own skills and career goals back to them, but with the structural rigour of a virtual supervisory team. It’s not about the AI "giving" the idea; it’s about the AI forcing the student to defend and refine their own concepts until they hold water. The Framework: 3 Months of Rigour This prompt is specifically designed for intensive/conversion MSc or summer capstone projects . It assumes a tight 12-week implementation window. By forcing the AI to w...
Post a Comment
Read more