InfoTech Assistant: A Multimodal Conversational Agent for InfoTechnology Web Portal Queries

Akhila Kudupudi; Charitha Chalasani; Duoduo Liao; Sai Surya Gadiraju; Santosh Kasula

REVIEW 3 major objections 6 minor 39 references

InfoTech Assistant: A Multimodal Conversational Agent for InfoTechnology Web Portal Queries

T0 review · 3 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read RAG chatbot answers bridge-tech lookup queries at ~95% accuracy.

desk verdict A standard RAG chatbot for one FHWA portal, with a headline accuracy claim that rests on a circular metric and an internally inconsistent evaluation table. read the letter →

arxiv 2412.16412 v2 pith:2WU5MSMX submitted 2024-12-21 cs.CL

Sai Surya Gadiraju , Duoduo Liao , Akhila Kudupudi , Santosh Kasula , Charitha Chalasani This is my paper · ORCID

classification cs.CL

keywords retrieval-augmentedgenerationbridgeinspectionmultimodalchatbotLlama3.1Mistral7Bcosinesimilaritywebscrapinginfrastructuretechnology

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The InfoTech Assistant is a pilot chatbot that answers lookup questions about bridge inspection and assessment technologies by retrieving scraped web-portal pages and letting an LLM generate the reply. The paper's central claim is that this retrieval-augmented system answers domain-specific queries correctly about 95 percent of the time with Llama 3.1 and about 93 percent with Mistral 7B, where 'correct' means the vectorized answer scores at least 0.85 cosine similarity to the expected response. The system also returns relevant images from the portal alongside text, and pairs a source-based 'bot' answer with a shorter LLM summary. If the claim holds, it shows that a modest, locally hosted model plus a small structured database can serve as a dependable reference tool for infrastructure professionals.

What carries the argument

The load-bearing mechanism is a retrieval-augmented generation (RAG) loop: a JSON database of scraped pages, an all-mpnet-base-v2 embedding model that retrieves the closest passage, and an LLM that generates the answer from that passage. The evaluation identity is Equation 3: accuracy is the share of test cases whose cosine similarity between the expected and actual response embeddings is at least 0.85. The same embedding model does double duty, both finding relevant context and scoring whether the answer matches.

What would settle it

Have a bridge-inspection expert independently score the 15 test answers as correct, incorrect, or partial, and compare those labels with the 0.85 cosine threshold; if many answers above the cutoff are judged wrong, or many below it are judged right, the accuracy metric is not measuring answer correctness.

Watch

Extended reading notes

Core claim

The paper reports that a RAG chatbot built on 41 scraped bridge-technology pages can handle conversational queries such as 'What is Electrical Resistivity?' and 'What are benefits of Hammer Sounding?' with high similarity to expected answers. Over 15 test questions, Llama 3.1 8B reaches an overall accuracy of 95 percent and Mistral 7B reaches 93 percent, with latency between roughly 10 and 22 seconds on a local workstation. The authors frame the contribution as a domain-specific, multimodal conversational agent: retrieval is grounded in the portal's own text, and images are pulled from the same source and displayed with the reply. They also note the dual-response format, where the bot response reproduces the source passage and the LLM response summarizes it, and they acknowledge that cosine similarity can underrate a factually correct answer that adds explanatory content.

Load-bearing premise

The reported accuracy rests on the assumption that a cosine similarity of 0.85 or higher, measured against a response drawn from the very passage the system retrieves, is a valid test of whether an answer is correct.

Editorial extensions

If this is right

A locally hosted 8B parameter model, running without cloud APIs, can reach roughly 95 percent accuracy on a narrow technical vocabulary drawn from the portal.
Llama 3.1 8B outperforms Mistral 7B on similarity-defined accuracy, while Mistral's latency is comparable or lower, so model choice becomes a speed-versus-accuracy trade-off.
The dual-response design means a professional can read the verbatim source passage for precision and the LLM summary for a quick overview.
The same scraping-to-JSON-to-RAG pipeline can be extended to the pavements, tunnels, and utilities sections already included in the dataset.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

In the editor's view, the 95 percent figure is best read as retrieval-and-paraphrase fidelity, not independent factual accuracy, because the expected response is the same scraped passage the retriever feeds the model.
A testable extension: run the same pipeline on a second technical portal and check whether the 0.85 cosine threshold keeps accuracy above 90 percent without retuning.
The evaluation is likely sensitive to question phrasing; a correct answer expressed differently from the source text could score below 0.85 even though a human would accept it.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

This is a standard RAG chatbot applied to one FHWA web portal, and the headline 95% accuracy figure rests on a metric that doesn't test correctness. What's genuinely here is a clear system description: they scraped 41 bridge-technology pages, stored the text and images in JSON, and built a Flask app with Llama 3.1 and Mistral 7B behind it. The appendix tables on Hammer Sounding and Magnetic Particle Testing give a concrete picture of the data. The latency measurements and model comparison are honest, and the authors acknowledge in Section V that the similarity metric has limitations.

The soft spots are in the evaluation. Equation 3 defines accuracy as the fraction of responses whose cosine similarity to the expected answer exceeds 0.85. The expected answer is the scraped source passage, and the RAG system retrieves that exact passage and feeds it to the LLM as context. So the similarity score measures how close the model's output is to its own input text, not whether the answer is factually correct for the user's query. The paper's own Section V admission—that Llama gave correct but more detailed answers with lower similarity—shows the metric does not track correctness.

Table VI is internally inconsistent with the stated method. Every reported similarity score is above the 0.85 threshold, so per the paper's own definition every question should be 100% accurate. Instead the table shows per-question accuracy values from 0.89 to 0.99. That doesn't match the binary correct/incorrect calculation in Equation 3. Either there were multiple unstated trials per question, or the accuracy column is doing something else; neither possibility is explained.

There's also no human evaluation, no held-out set, no error bars, and no released code or data. Fifteen questions is too small for a meaningful percentage claim. The citation list is sloppy: the sentence-transformer model is attributed to the wrong paper, and several references are off-topic.

Net: this is a pilot demo, not a research contribution. The system could be a useful internal tool for bridge inspectors, and the paper is an honest write-up of a standard pipeline. But the central accuracy claim is not supported, and the evaluation is self-referential. I would not send this to serious peer review in its current form. If the authors redo the evaluation with human judgments or an independent QA benchmark, it might become a small, publishable systems paper. As it stands, a desk reject is the right call.

Referee Report

3 major / 6 minor

Summary. The paper describes a pilot multimodal conversational agent, 'InfoTech Assistant,' for the FHWA InfoTechnology web portal. The system scrapes 41 bridge-technology pages, stores text and images in JSON, and uses a Flask back end with either Llama 3.1 8B or Mistral-7B-Instruct-v0.2 together with sentence-embedding-based retrieval (all-mpnet-base-v2) to answer domain-specific questions. The central reported result is that the assistant achieves about 95% (Llama 3.1) and 93% (Mistral 7B) accuracy on 15 bridge-technology queries, where correctness is defined by a cosine similarity of at least 0.85 between the generated response and an 'expected' response (Section IV-D-3, Eq. 3). The paper also reports latency ranges, sample interactions, and a qualitative discussion of retrieval limitations.

Significance. If the reported accuracy were valid, the work would be a modest but useful demonstration of a RAG-based domain chatbot for infrastructure professionals. The system integration (web scraping, JSON storage, embedding-based retrieval, local LLM serving) is clearly described and the appendix provides concrete sample data. However, the central quantitative claim is not supported: the evaluation metric is self-referential, comparing generated answers against the very source passages the RAG pipeline retrieves and supplies to the LLM. There is no human evaluation, no independent ground truth, no held-out test set, and no error analysis beyond a brief acknowledgment in Section V that cosine similarity 'may not be the most reliable metric.' The paper's own stated limitation, together with the internal inconsistency of Table VI, makes the headline accuracy figures unsubstantiated. The contribution is therefore reduced to a system-description pilot, not a validated evaluation of response quality.

major comments (3)

[Section IV-D-3, Eq. (3)] The accuracy definition is circular. 'Expected' responses are evidently the scraped FHWA source passages stored in the JSON database (as shown in Tables III and IV), and the RAG system retrieves those exact passages before the LLM summarizes or paraphrases them into the 'actual' response. Cosine similarity between the generated answer and the source text it was constructed from measures lexical-semantic overlap, not factual correctness for the user's question. No external labels, human ratings, or independent benchmark are used. The paper itself concedes in Section V that in some cases Llama produced 'correct and detailed answers' with lower similarity, which directly undercuts the claim that the 0.85 threshold tracks correctness. The 95%/93% accuracy figures therefore do not establish the assistant's reliability.
[Table VI and Section IV-D-3] The per-question accuracy values are internally inconsistent with the stated definition. Equation 3 defines accuracy as the percentage of correct responses among total test cases, and Section IV-D-3 states that a response is correct if its cosine similarity is 0.85 or higher. For a single test case this yields exactly 100% or 0%, yet Table VI reports accuracies such as 0.91, 0.93, and 0.99. If there were multiple repeated trials per question, the paper does not say so, and the accuracy values are suspiciously close to the corresponding similarity scores, suggesting either an informal mapping or unstated aggregation. This inconsistency makes the quantitative evaluation impossible to interpret and requires correction or explicit clarification of the testing protocol.
[Section IV-D and Section IV-D-2] The evaluation lacks statistical rigor for the claims made. Only 15 questions are used, with no description of how they were selected, whether they cover the full set of 41 technologies, or whether they are a held-out set. Despite the text in Section IV claiming testing with 'technical and non-technical users' and 'multiple testing rounds,' no user study protocol, participant numbers, or satisfaction measurements are reported. Latency ranges (15-20 seconds for Llama, 10-22 for Mistral) are given without error bars or condition details beyond the single hardware configuration in Table V. The absence of these details, combined with the metric problems above, means the paper's evaluative claims are not reproducible from the information provided.

minor comments (6)

[Section I] There is a typo in the introduction: 'to to enhance information retrieval' should read 'to enhance information retrieval.'
[Section III-E-3, Eq. (1)] Equation (1) as printed is incomplete: the denominator is written as 'P exp(zj/T )' instead of a sum over j of exp(zj/T), making the softmax formula incorrect as typeset.
[Section IV-D-3] The text states that cosine similarity values 'calculated using Equation 3,' but Equation 3 is the accuracy formula; the cosine similarity is defined in Equation 2. The cross-reference should be corrected.
[Section III-C and Reference [5]] The discussion of all-mpnet-base-v2 cites reference [5] (Xiao, 2022), but that reference appears to describe a different QA chatbot model, not the Microsoft sentence-transformer model. The citation should be replaced with the actual all-mpnet-base-v2 source.
[Acknowledgment] The acknowledgment refers to 'the LLaMA model' and thanks Meta and Mistral; for consistency with the rest of the paper, the Llama 3.1 model name should be used, and formatting inconsistencies such as 'FHW A' should be cleaned up.
[Table VI] The table would benefit from a column explaining the number of trials or repetitions per question, since the reported accuracy values appear to be fractional rather than binary; without this information the table is misleading.

Circularity Check

1 steps flagged · score 7.0 of 10

Headline accuracy is scored against the very source text the RAG system retrieves and summarizes, so the 95% figure is largely self-referential.

self definitional [Section IV-D-3 (Similarity Calculation and Accuracy Evaluation), Eq. 3; Section IV-C (Comparison of Bot and LLM Responses)]
""Expected and actual responses were vectorized using the Sentence-Transformer model [5], and accuracy was determined as the percentage of correct responses among the total test cases." (IV-D-3) "The Bot response is directly generated from data scraped from the official InfoTechnology website... the LLM response is a summarized version of the same information." (IV-C)"

The 'expected' response is the scraped InfoTechnology/FHWA passage stored in the JSON database, and the RAG pipeline retrieves exactly that passage for the LLM to summarize as the 'actual' response. Accuracy (Eq. 3) is therefore defined as whether the generated answer's cosine similarity (Eq. 2) to its own source text exceeds 0.85. This measures paraphrase/lexical overlap rather than factual correctness, so a competent RAG system will score high even when it hallucinates details, omits constraints, or answers a different question. No human labels or external QA benchmark validate the threshold; the paper itself concedes in Section V that correct, more detailed answers can receive lower similarity, confirming the metric does not track correctness.

full rationale

The central accuracy claim reduces to a self-referential comparison. Section IV-C states the LLM response is 'a summarized version of the same information' as the Bot response 'directly generated from data scraped from the official InfoTechnology website.' Section IV-D-3 then defines correctness as cosine similarity of at least 0.85 between 'expected' and 'actual' vectorized responses, with no independent ground truth. Because the RAG system retrieves the source passage that serves as the expected answer, high similarity is assured by pipeline construction; it is not evidence of domain accuracy. The paper's own conclusion (Section V) that similarity 'may not be the most reliable metric' and that correct detailed answers can score lower further undermines the metric. This is not a self-citation or imported-uniqueness issue; it is a definitional/evaluation circularity. The system itself may work, but the paper's headline quantitative claim is not supported by the reported experiment. Score 7 rather than 8 because the circularity is in the evaluation design rather than in a self-citation chain, and there is no indication of deliberate result forcing; nonetheless the main claim reduces by construction.

Assumptions & free parameters 2 free parameters · 3 assumptions · 1 invented entities

The central claim rests on domain assumptions about the scraped data being ground truth, cosine similarity being a correctness oracle, and 15 self-authored questions being representative. The hand-set threshold and temperature are the only free parameters; no predictive model is fitted, but the evaluation is anchored to the source text that the RAG system retrieves.

free parameters (2)

cosine similarity correctness threshold = 0.85
Equation 3 defines correct as similarity >= 0.85; the threshold is chosen ad hoc with no human validation, and the reported accuracy numbers depend directly on it.
LLM temperature = 0.7
Section III-E-3 sets temperature to 0.7 to balance determinism and diversity; it is a hand-chosen hyperparameter, not fitted to data, but it affects response variability.

assumptions (3)

domain assumption Scraped FHWA InfoTechnology web pages are an accurate and complete ground truth for bridge inspection technology answers.
The system and evaluation treat scraped text as the expected response; Section III-B describes data collection and the appendix shows sample source records.
domain assumption Cosine similarity between the embedding of the expected response and the generated response is a valid proxy for answer correctness.
Section IV-D-3 applies a 0.85 threshold to define accuracy, but no correlation with human judgments or an external QA benchmark is established.
domain assumption The 15 test questions and their expected answers are representative of real user queries in bridge technology.
Section IV-D states 15 different questions were used, but the selection process and coverage analysis are not described, and no external test set is used.

invented entities (1)

None
purpose: No new theoretical entities are introduced.
The InfoTech Assistant is an assembled software system; it does not postulate new particles, forces, dimensions, or conserved quantities.

how reviews work

0 comments

Cite this review

Pith. "Pith review of InfoTech Assistant: A Multimodal Conversational Agent for InfoTechnology Web Portal Queries." pith.science (2026). https://pith.science/paper/2WU5MSMX

@misc{pith2026241216412,
  author       = {Pith},
  title        = {Pith review of: InfoTech Assistant: A Multimodal Conversational Agent for InfoTechnology Web Portal Queries},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2WU5MSMX}},
  note         = {Machine review of arXiv:2412.16412}
}

read the original abstract

This pilot study presents the development of the InfoTech Assistant, a domain-specific, multimodal chatbot engineered to address queries in bridge evaluation and infrastructure technology. By integrating web data scraping, large language models (LLMs), and Retrieval-Augmented Generation (RAG), the InfoTech Assistant provides accurate and contextually relevant responses. Data, including textual descriptions and images, are sourced from publicly available documents on the InfoTechnology website and organized in JSON format to facilitate efficient querying. The architecture of the system includes an HTML-based interface and a Flask back end connected to the Llama 3.1 model via LLM Studio. Evaluation results show approximately 95 percent accuracy on domain-specific tasks, with high similarity scores confirming the quality of response matching. This RAG-enhanced setup enables the InfoTech Assistant to handle complex, multimodal queries, offering both textual and visual information in its responses. The InfoTech Assistant demonstrates strong potential as a dependable tool for infrastructure professionals, delivering high accuracy and relevance in its domain-specific outputs.

Figures

Figures reproduced from arXiv: 2412.16412 by the authors.

**Figure 2.** InfoTechnology Web Portal UI with InfoTech Assistant [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 3.** User Queries and Responses by the InfoTech Assistant. [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

Figures from the paper (1 more)

**Figure 4.** Figure 4: Similarity scores and overall accuracies of Llama 3.1 [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 23 canonical work pages

[1]

”FHW A InfoTechnology Platform: Supporting Decision-Making for Infrastructure Management”

Federal Highway Administration. ”FHW A InfoTechnology Platform: Supporting Decision-Making for Infrastructure Management”. U.S. De- partment of Transportation. 2020

work page 2020
[2]

H. S. Park and B. D. Smith. ”Centralized Information Systems in Infrastructure Management: A Case Study on the FHW A InfoTech Platform”. International Journal of Civil Engineering, 2017, 15(6), 495- 503

work page 2017
[3]

FHW A InfoTechnology,

Federal Highway Administration. “FHW A InfoTechnology,” Dot.gov. https://infotechnology.fhwa.dot.gov/ (accessed Nov. 15, 2024)

work page 2024
[4]

Bridge FHW A InfoTechnology,

FHW A InfoTechnology. “Bridge FHW A InfoTechnology,” Dot.gov. https://infotechnology.fhwa.dot.gov/bridge/. (Accessed: Nov. 15, 2024)

work page 2024
[5]

Xiao, Yunze. (2022). A Transformer-based Attention Flow Model for Intelligent Question and Answering Chatbot. 167-170. 10.1109/IC- CRD54409.2022.9730454.(accessed: Nov. 12, 2024)

arXiv 2022
[6]

LM Studio - Experiment with Local LLMs

LM Studio. “LM Studio - Experiment with Local LLMs.” https:// lmstudio.ai/

work page
[7]

P. P. Ghadekar, S. Mohite, O. More, P. Patil, Sayantika, and S. Mangrule, ”Sentence Meaning Similarity Detector Using FAISS,” 2023 7th International Conference On Computing, Communication, Con- trol And Automation (ICCUBEA) , Pune, India, 2023, pp. 1–6. doi: 10.1109/ICCUBEA58933.2023.10392009.(accessed: Nov. 15, 2024)

arXiv 2023
[8]

GitHub - alfredodeza/learn-retrieval-augmented- generation: Examples and demos on how to use Retrieval Augmented Generation with Large Language Models

Alfredodeza. “GitHub - alfredodeza/learn-retrieval-augmented- generation: Examples and demos on how to use Retrieval Augmented Generation with Large Language Models.” GitHub.https://github.com/ alfredodeza/learn-retrieval-augmented-generation.(accessed: Nov. 01, 2024)

work page 2024

Show all 39 references

[9]

Anghelescu, P., & Nicolaescu, S. V . (2018, June). Chatbot application using search engines and teaching methods. In 2018 10th international conference on electronics, computers and artificial intelligence (ECAI) (pp. 1-6). IEEE

2018
[10]

S., Fonseca-Delgado, R., Reyes-Chac ´on, I., Vizcaino- Imacana, P., & Morocho-Cayamcela, M

Cabezas, D. S., Fonseca-Delgado, R., Reyes-Chac ´on, I., Vizcaino- Imacana, P., & Morocho-Cayamcela, M. E. Integrating a LLaMa-based Chatbot with Augmented Retrieval Generation as a Complementary Educational Tool for High School and College Students.(accessed: Nov. 01, 2024)

2024
[11]

(2023, May)

Singh, V ., Rohith, Y ., Prakash, B., & Kumari, U. (2023, May). ChatBot using Python Flask. In 2023 7th International Conference on Intel- ligent Computing and Control Systems (ICICCS) (pp. 1182-1185). IEEE.(accessed: Nov. 01, 2024)

2023
[12]

U., Cirik, V ., & Cho, K

Dunn, M., Sagun, L., Higgins, M., Guney, V . U., Cirik, V ., & Cho, K. (2017). Searchqa: A new q&a dataset augmented with context from a search engine. arXiv preprint arXiv:1704.05179.(accessed: Nov. 15, 2024.)

2017 arXiv
[13]

Li, Y ., Li, W., & Nie, L. (2022). Dynamic graph reasoning for con- versational open-domain question answering. ACM Transactions on Information Systems (TOIS), 40(4), 1-24

2022
[14]

Dimitriadis, D. (2023). Machine learning and natural language process- ing techniques for question answering (Doctoral dissertation, ARISTO- TLE UNIVERSITY OF THESSALONIKI)

2023
[15]

Jack” McGowan,

John J. “Jack” McGowan, ”Chapter 14 Project Haystack Data Stan- dards,” in Energy and Analytics BIG DATA and Building Technology Integration , River Publishers, 2015, pp.237-243

2015
[16]

Meet Sidekick, Skanska’s new AI chatbot,

M. Thibault, “Meet Sidekick, Skanska’s new AI chatbot,” Construction Dive, Mar. 06, 2024. [Online]. Available: https://www.constructiondive.com/news/sidekick-skanskas-ai- construction-chatbot/709350/.(Accessed: Nov. 10, 2024)

2024
[17]

& Wang, T

Bajaj, P., Campos, D., Craswell, N., Deng, L., Gao, J., Liu, X., ... & Wang, T. (2016). Ms marco: A human generated machine reading comprehension dataset. arXiv preprint arXiv:1611.09268

2016 arXiv
[18]

L., & Parikh, D

Antol, S., Agrawal, A., Lu, J., Mitchell, M., Batra, D., Zitnick, C. L., & Parikh, D. (2015). Vqa: Visual question answering. In Proceedings of the IEEE international conference on computer vision (pp. 2425-2433)

2015
[19]

H. K. Skrodelis, A. Romanovs, N. Zenina and H. Gorskis, ”The Latest in Natural Language Generation: Trends, Tools and Applications in In- dustry,” 2023 IEEE 10th Jubilee Workshop on Advances in Information, Electronic and Electrical Engineering (AIEEE), Vilnius, Lithuania, 202...

2023
[20]

Diagramming Powered by Intelligence

LucidChart. “Diagramming Powered by Intelligence.” Available: https: //www.lucidchart.com/.(accessed: Nov. 01, 2024)

2024
[21]

Stack Overflow. ”Using Selenium to click page and scrape info from routed page.” Available: https://stackoverflow.com/questions/71786531/ using-selenium-to-click-page-and-scrape-info-from-routed-page. Accessed: Nov. 02, 2024

2024
[22]

(2020, May)

Thosani, P., Sinkar, M., Vaghasiya, J., & Shankarmani, R. (2020, May). A self learning chat-bot from user interactions and preferences. In 2020 4th International Conference on Intelligent Computing and Control Systems (ICICCS) (pp. 224-229). IEEE. 10.1109/ICICCS48265.2020. 9120912

2020
[23]

M. G, M. M, S. R and I. Ritharson P, ”Chatbots Embracing Artifi- cial Intelligence Solutions to Assist Institutions in Improving Student Interactions,” 2023 International Conference on Circuit Power and Computing Technologies (ICCPCT), Kollam, India, 2023, pp. 912-916, doi: 10...

2023
[24]

Omrani, P., Hosseini, A., Hooshanfar, K., Ebrahimian, Z., Toosi, R., & Akhaee, M. A. (2024, April). Hybrid Retrieval-Augmented Generation Approach for LLMs Query Response Enhancement. In 2024 10th International Conference on Web Research (ICWR) (pp. 22-26). IEEE. doi: 10.1109/...

2024
[25]

ICIS 2024 Proceedings

Zeng, Jicheng; Liu, Xiaochen; and Fang, Yulin, ”Influence of Leaderboard and Trial Space on Large Language Models Popularity” (2024). ICIS 2024 Proceedings. 9. https://aisel.aisnet.org/icis2024/digital emergsoc/digital emergsoc/9

2024
[26]

Q., Sablayrolles, A., Mensch, A., Bamford, C., Chaplot, D

Jiang, A. Q., Sablayrolles, A., Mensch, A., Bamford, C., Chaplot, D. S., Casas, D. D. L., ... & Sayed, W. E. (2023). Mistral 7B. arXiv preprint arXiv:2310.06825

2023 arXiv
[27]

Ait-Mlouk, A., & Jiang, L. (2020). KBot: a Knowledge graph based chatBot for natural language understanding over linked data. IEEE Access, 8, 149220-149230. doi: 10.1109/ACCESS.2020.3011848

2020
[28]

He, Z., Shu, W., Ge, X., et al. (2024). Llama Scope: Extracting Millions of Features from Llama-3.1-8B with Sparse Autoencoders. arXiv preprint arXiv:2410.20526

2024 arXiv
[29]

Bengio, Y ., Goodfellow, I., & Courville, A. (2017). Deep learning (V ol. 1). Cambridge, MA, USA: MIT press.(Accessed: Nov. 14, 2024)

2017
[30]

(2024, May)

Perkovi ´c, G., Drobnjak, A., & Boti ˇcki, I. (2024, May). Hallucinations in llms: Understanding and addressing challenges. In 2024 47th MIPRO ICT and Electronics Convention (MIPRO) (pp. 2084-2088). IEEE. doi: 10.1109/MIPRO60963.2024.10569238

2024
[31]

(2024, September)

Rosati, R., Antonini, F., Muralikrishna, N., Tonetto, F., & Mancini, A. (2024, September). Improving Industrial Question Answering Chatbots with Domain-Specific LLMs Fine-Tuning. In 2024 20th IEEE/ASME International Conference on Mechatronic and Embedded Systems and Applicatio...

2024
[32]

Bhattacharjee, A

S. Bhattacharjee, A. Das, U. Bhattacharya, S. K. Parui and S. Roy, ”Sen- timent analysis using cosine similarity measure,” 2015 IEEE 2nd Inter- national Conference on Recent Trends in Information Systems (ReTIS), Kolkata, India, 2015, pp. 27-32, doi: 10.1109/ReTIS.2015.7232847

2015
[33]

M., Embar, V ., & Raghunathan, K

Jayanthi, S. M., Embar, V ., & Raghunathan, K. (2021). Evaluating pretrained transformer models for entity linking in task-oriented dialog. arXiv preprint arXiv:2112.08327

2021 arXiv
[34]

Peeperkorn, M., Kouwenhoven, T., Brown, D., & Jordanous, A. (2024). Is temperature the creativity parameter of large language models?. arXiv preprint arXiv:2405.00492

2024 arXiv
[35]

& Zhou, J

Dong, G., Yuan, H., Lu, K., Li, C., Xue, M., Liu, D., ... & Zhou, J. (2023). How abilities in large language models are affected by super- vised fine-tuning data composition. arXiv preprint arXiv:2310.05492

2023 arXiv
[36]

P., Zhang, Y ., Briggs, J., Blake, C., Levy-Kramer, J., Bal- anca, P.,

Perez, S. P., Zhang, Y ., Briggs, J., Blake, C., Levy-Kramer, J., Bal- anca, P., ... & Fitzgibbon, A. W. (2023). Training and inference of large language models using 8-bit floating point. arXiv preprint arXiv:2309.17224

2023 arXiv
[37]

(2024, July)

D’Urso, S., Martini, B., & Sciarrone, F. (2024, July). A Novel LLM Architecture for Intelligent System Configuration. In 2024 28th Interna- tional Conference Information Visualisation (IV) (pp. 326-331). IEEE

2024
[38]

Vrana, J., Meyendorf, N., Ida, N., & Singh, R. (2022). Introduction to NDE 4.0. In: Meyendorf, N., Ida, N., Singh, R., & Vrana, J. (eds) Handbook of Nondestructive Evaluation 4.0. Springer, Cham. Retrieved from https://doi.org/10.1007/978-3-030-73206-6 \ 43

2022 doi
[39]

dry powder

Almazrouei, E., Alobeidli, H., Alshamsi, A., Cappelli, A., Cojocaru, R., Debbah, M., ... & Penedo, G. (2023). The falcon series of open language models. arXiv preprint arXiv:2311.16867. APPENDIX Sample Data from InfoTech Assistant JSON file TABLE III: Supplementary Data Table:...

2023 arXiv

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.

[1] [1]

”FHW A InfoTechnology Platform: Supporting Decision-Making for Infrastructure Management”

Federal Highway Administration. ”FHW A InfoTechnology Platform: Supporting Decision-Making for Infrastructure Management”. U.S. De- partment of Transportation. 2020

work page 2020

[2] [2]

H. S. Park and B. D. Smith. ”Centralized Information Systems in Infrastructure Management: A Case Study on the FHW A InfoTech Platform”. International Journal of Civil Engineering, 2017, 15(6), 495- 503

work page 2017

[3] [3]

FHW A InfoTechnology,

Federal Highway Administration. “FHW A InfoTechnology,” Dot.gov. https://infotechnology.fhwa.dot.gov/ (accessed Nov. 15, 2024)

work page 2024

[4] [4]

Bridge FHW A InfoTechnology,

FHW A InfoTechnology. “Bridge FHW A InfoTechnology,” Dot.gov. https://infotechnology.fhwa.dot.gov/bridge/. (Accessed: Nov. 15, 2024)

work page 2024

[5] [5]

Xiao, Yunze. (2022). A Transformer-based Attention Flow Model for Intelligent Question and Answering Chatbot. 167-170. 10.1109/IC- CRD54409.2022.9730454.(accessed: Nov. 12, 2024)

arXiv 2022

[6] [6]

LM Studio - Experiment with Local LLMs

LM Studio. “LM Studio - Experiment with Local LLMs.” https:// lmstudio.ai/

work page

[7] [7]

P. P. Ghadekar, S. Mohite, O. More, P. Patil, Sayantika, and S. Mangrule, ”Sentence Meaning Similarity Detector Using FAISS,” 2023 7th International Conference On Computing, Communication, Con- trol And Automation (ICCUBEA) , Pune, India, 2023, pp. 1–6. doi: 10.1109/ICCUBEA58933.2023.10392009.(accessed: Nov. 15, 2024)

arXiv 2023

[8] [8]

GitHub - alfredodeza/learn-retrieval-augmented- generation: Examples and demos on how to use Retrieval Augmented Generation with Large Language Models

Alfredodeza. “GitHub - alfredodeza/learn-retrieval-augmented- generation: Examples and demos on how to use Retrieval Augmented Generation with Large Language Models.” GitHub.https://github.com/ alfredodeza/learn-retrieval-augmented-generation.(accessed: Nov. 01, 2024)

work page 2024