slides · global inventory
Triple Take
Arc collection for the active corpus filters, with each row opening the slide detail view.
Matching slides 13,579
Current window 9,541-9,570
Pages 453
Rows per page 30
Arc
9,541-9,570 of 13,579
108
The slide uses a numbered list format to present forward-looking statements.2022 Air Street Capital The State of AI Report 2022 · Air Street Capital
Consulting deck VC research Beat · The Action (Now What)Loop · Time Machine1 hit
109
2022 Air Street Capital The State of AI Report 20222022 Air Street Capital The State of AI Report 2022 · Air Street Capital
Consulting deck VC research Beat · The Action (Now What)1 hit
11
2023 Air Street Capital The State of AI Report 20232023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)1 hit
12
The chart shows a comparison of GPT-4, GPT-4 (no vision), and GPT-3.5 across various academic and professional exams.2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Pattern Hunter1 hit
13
Includes a process diagram illustrating the RLHF training pipeline.2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Pattern Hunter1 hit
14
The slide references Berkeley research on imitation learning and Meta's LLaMA-2 paper findings on RLHF.2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Pattern Hunter1 hit
15
The slide discusses alternatives to Reinforcement Learning from Human Feedback (RLHF) in AI training.2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Pattern Hunter1 hit
16
Includes a screenshot of a tweet by William Falcon mocking the lack of detail in the GPT-4 technical report.2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Pattern Hunter1 hit
17
The slide discusses the release of LLaMa, its performance metrics (RACE benchmark), and the controversy regarding its 'open-source' status.2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Pattern Hunter1 hit
18
The slide discusses the ecosystem of open-source LLMs, mentioning specific models like MPT-30B, Falcon, RedPajama, Pythia, and Vicuna, and techniques like LoRA.2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Pattern Hunter1 hit
19
The slide uses a stacked bar chart to show human evaluation of LLaMa-2 helpfulness against other models.2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Pattern Hunter1 hit
20
The chart maps model release/popularity over time (x-axis) against MMLU performance (y-axis), with bubble size representing social media mentions.2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Pattern Hunter1 hit
21
The chart tracks the share of different research topics in the AI field.2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Pattern Hunter1 hit
22
The slide presents a critique of the 'emergent capabilities' phenomenon in LLMs, citing Stanford research that suggests these are artifacts of metric choice.2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Zoom In1 hit
23
The slide discusses technical methods like FlashAttention, ALiBi, RoPE, and Positional Interpolation in the context of LLM scaling.2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Zoom In1 hit
24
The slide references the 'Lost in the Middle' research paper by Samaya.ai, UC Berkeley, Stanford, and LMSYS.org.2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Zoom In1 hit
25
Part of the State of AI 2023 report.2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Zoom In1 hit
26
The slide highlights the shift from 'more data/parameters' to 'better data' in LLM training, specifically referencing the 'phi' model series.2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Zoom In1 hit
27
The slide uses a visual comparison (meme vs. truth) to explain the shift from dense models to Mixture of Experts (MoE) architectures.2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Zoom In1 hit
28
The slide uses three separate line charts to illustrate data exhaustion timelines for different data types.2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Zoom In1 hit
29
The chart shows Top-1 Accuracy vs Parameters (M) for models trained on 'Real' vs 'Real + Generated' data, indicating performance gains.2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Zoom In1 hit
30
Includes logos of University of Maryland, Google, Google DeepMind, ETH Zurich, and Princeton.2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Zoom In1 hit
31
Discusses the trade-off between training epochs, model size, and overfitting.2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Zoom In1 hit
32
The slide contrasts structured quantitative evaluation (HELM/Hugging Face) with qualitative user sentiment ('vibes').2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Pattern Hunter1 hit
33
Includes a process diagram showing the training pipeline for Code Llama models.2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Pattern Hunter1 hit
34
The slide uses two process diagrams to explain how AlphaDev functions as an RL agent.2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Pattern Hunter1 hit
35
The slide compares different prompting architectures (IO, CoT, CoT-SC, ToT) visually.2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Pattern Hunter1 hit
36
The slide uses bar charts to compare accuracy, verbosity, and mismatch metrics for two model versions over time.2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Pattern Hunter1 hit
37
Includes a small example box illustrating tool usage (QA and Calculator).2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Pattern Hunter1 hit
38
The chart shows the 'Minecraft Tech Tree' progression of different agents, with Voyager (ours) showing the highest performance.2023 Air Street Capital The State of AI Report 2023 · Air Street Capital
Consulting deck VC research Beat · The Facts (What)Loop · Pattern Hunter1 hit