Human Leaderboard: overall

  • Ranking: The position of the model in the leaderboard as ordered by Overall Score
  • Organization: The group responsible for the model or forecasts
  • Model: The LLM model & prompt info or the human group and forecast aggregation method
    • zero shot: used a zero-shot prompt
    • scratchpad: used a scratchpad prompt with instructions that outline a procedure the model should use to reason about the question
    • with freeze values: means that, for questions from market sources, the prompt was supplemented with the aggregate human forecast from the relevant platform on the day the question set was generated
    • with news: means that the prompt was supplemented with relevant news summaries obtained through an automated process
  • Dataset Score: The average Brier score across all questions sourced from datasets
  • Market Score (resolved): The average Brier score across all resolved questions sourced from prediction markets and forecast aggregation platforms
  • Market Score (unresolved): The average Brier score across all unresolved questions sourced from prediction markets and forecast aggregation platforms
  • Market Score (overall): The average Brier score across all questions sourced from prediction markets and forecast aggregation platforms
  • Overall Resolved Score: The average of the Dataset Score and the Market Score (resolved) columns
  • Overall Score: The average of the Dataset Score and the Market Score (overall) columns
  • Overall Score 95% CI: The 95% confidence interval for the Overall Score
  • Pairwise p-value comparing to No. 1 (bootstrapped): The p-value calculated by bootstrapping the differences in overall score between each model and the best forecaster (the group with rank 1) under the null hypothesis that there's no difference.
  • Pct. more accurate than No. 1: The percent of questions where this forecaster had a better overall score than the best forecaster (with rank 1)
  • Pct. imputed: The percent of questions for which this forecaster did not provide a forecast and hence had a forecast value imputed (0.5 for dataset questions and the aggregate human forecast on the forecast due date for questions sourced from prediction markets or forecast aggregation platforms)
Ranking Organization Model Dataset Score (N=210) Market Score (resolved) (N=14) Market Score (unresolved) (N=63) Market Score (overall) (N=77) Overall Resolved Score (N=224) Overall Score (N=287) Overall Score 95% CI Pairwise p-value comparing to No. 1 (bootstrapped) Pct. more accurate than No. 1 Pct. Imputed
1 ForecastBench Superforecaster median forecast 0.135 0.119 0.036 0.051 0.127 0.093 [0.073, 0.112] 0% 0%
2 ForecastBench Public median forecast 0.167 0.165 0.022 0.048 0.166 0.107 [0.09, 0.125] 0.009 30% 0%
3 Anthropic Claude-3-5-Sonnet-20240620 (scratchpad with freeze values) 0.155 0.196 0.039 0.067 0.176 0.111 [0.09, 0.132] 0.01 33% 0%
4 OpenAI GPT-4 (zero shot with freeze values) 0.169 0.184 0.028 0.056 0.177 0.113 [0.093, 0.132] 0.003 37% 0%
5 Anthropic Claude-3-5-Sonnet-20240620 (zero shot with freeze values) 0.155 0.288 0.026 0.074 0.222 0.114 [0.09, 0.139] 0.008 36% 0%
6 Anthropic Claude-3-5-Sonnet-20240620 (scratchpad with news with freeze values) 0.160 0.215 0.038 0.070 0.188 0.115 [0.095, 0.135] 0.004 31% 0%
7 OpenAI GPT-4o (scratchpad with freeze values) 0.173 0.170 0.034 0.059 0.171 0.116 [0.098, 0.133] 0.002 30% 0%
8 OpenAI GPT-4-Turbo-2024-04-09 (zero shot with freeze values) 0.177 0.184 0.026 0.055 0.180 0.116 [0.097, 0.135] <0.001 37% 0%
9 OpenAI GPT-4o (scratchpad with news with freeze values) 0.179 0.119 0.045 0.059 0.149 0.119 [0.101, 0.137] <0.001 28% 0%
10 Mistral AI Mistral-Large-Latest (scratchpad with freeze values) 0.163 0.158 0.064 0.081 0.161 0.122 [0.107, 0.137] <0.001 24% 0%
11 Mistral AI Mistral-Large-Latest (zero shot with freeze values) 0.173 0.196 0.047 0.074 0.184 0.123 [0.102, 0.145] <0.001 28% 0%
12 Anthropic Claude-3-5-Sonnet-20240620 (scratchpad) 0.155 0.252 0.058 0.093 0.204 0.124 [0.103, 0.145] <0.001 29% 0%
13 Meta Llama-3-70b-Chat-Hf (zero shot with freeze values) 0.175 0.177 0.053 0.076 0.176 0.125 [0.104, 0.146] <0.001 30% 0%
14 Anthropic Claude-3-5-Sonnet-20240620 (superforecaster with news 3) 0.171 0.186 0.058 0.082 0.179 0.126 [0.108, 0.145] <0.001 28% 2%
15 Anthropic Claude-3-5-Sonnet-20240620 (superforecaster with news 1) 0.167 0.228 0.058 0.089 0.197 0.128 [0.107, 0.148] <0.001 28% 0%
16 Anthropic Claude-3-Opus-20240229 (zero shot with freeze values) 0.185 0.206 0.041 0.071 0.196 0.128 [0.106, 0.151] <0.001 33% 0%
17 OpenAI GPT-4 (scratchpad with freeze values) 0.188 0.173 0.046 0.069 0.181 0.128 [0.114, 0.143] <0.001 24% 0%
18 Anthropic Claude-3-5-Sonnet-20240620 (scratchpad with news) 0.160 0.222 0.069 0.097 0.191 0.128 [0.108, 0.149] <0.001 26% 0%
19 OpenAI GPT-4o (scratchpad) 0.173 0.199 0.059 0.084 0.186 0.128 [0.111, 0.146] <0.001 26% 0%
20 Anthropic Claude-2.1 (scratchpad with freeze values) 0.206 0.061 0.054 0.055 0.134 0.131 [0.116, 0.145] <0.001 30% 23%
21 Qwen Qwen1.5-110B-Chat (zero shot with freeze values) 0.196 0.170 0.042 0.065 0.183 0.131 [0.113, 0.149] <0.001 29% 0%
22 Meta Llama-3-70b-Chat-Hf (scratchpad with freeze values) 0.188 0.149 0.059 0.076 0.168 0.132 [0.117, 0.147] <0.001 24% 0%
23 Google Gemini-1.5-Pro (zero shot with freeze values) 0.190 0.201 0.048 0.076 0.196 0.133 [0.109, 0.157] <0.001 31% 0%
24 OpenAI GPT-4o (zero shot with freeze values) 0.205 0.185 0.036 0.063 0.195 0.134 [0.112, 0.156] <0.001 35% 0%
25 Anthropic Claude-3-Opus-20240229 (scratchpad with freeze values) 0.176 0.173 0.074 0.092 0.175 0.134 [0.115, 0.152] <0.001 26% 0%
26 OpenAI GPT-4o (scratchpad with news) 0.179 0.195 0.065 0.089 0.187 0.134 [0.113, 0.155] <0.001 25% 0%
27 Anthropic Claude-3-5-Sonnet-20240620 (zero shot) 0.155 0.308 0.071 0.114 0.231 0.134 [0.109, 0.16] <0.001 28% 0%
28 Google Gemini-1.5-Flash (scratchpad with freeze values) 0.195 0.164 0.058 0.077 0.179 0.136 [0.117, 0.155] <0.001 25% 0%
29 Google Gemini-1.5-Flash (zero shot with freeze values) 0.183 0.205 0.063 0.089 0.194 0.136 [0.11, 0.163] <0.001 30% 0%
30 OpenAI GPT-4-Turbo-2024-04-09 (zero shot) 0.177 0.235 0.067 0.098 0.206 0.137 [0.118, 0.156] <0.001 22% 0%
31 Mistral AI Mixtral-8x22B-Instruct-V0.1 (zero shot with freeze values) 0.195 0.183 0.056 0.079 0.189 0.137 [0.114, 0.16] <0.001 32% 0%
32 Anthropic Claude-3-Opus-20240229 (scratchpad) 0.176 0.189 0.080 0.100 0.182 0.138 [0.121, 0.155] <0.001 24% 0%
33 OpenAI GPT-4-Turbo-2024-04-09 (scratchpad with freeze values) 0.191 0.152 0.070 0.085 0.171 0.138 [0.116, 0.16] <0.001 27% 0%
34 Mistral AI Mixtral-8x22B-Instruct-V0.1 (scratchpad with freeze values) 0.190 0.187 0.063 0.086 0.189 0.138 [0.122, 0.154] <0.001 25% 0%
35 Anthropic Claude-2.1 (scratchpad) 0.206 0.111 0.061 0.070 0.159 0.138 [0.121, 0.156] <0.001 28% 23%
36 Google Gemini-1.5-Pro (scratchpad with news with freeze values) 0.181 0.176 0.079 0.096 0.179 0.138 [0.12, 0.157] <0.001 25% 0%
37 Google Gemini-1.5-Pro (scratchpad with freeze values) 0.179 0.222 0.072 0.099 0.200 0.139 [0.121, 0.157] <0.001 26% 0%
38 OpenAI GPT-4 (scratchpad) 0.188 0.176 0.079 0.096 0.182 0.142 [0.128, 0.157] <0.001 20% 0%
39 Mistral AI Mistral-Large-Latest (scratchpad) 0.163 0.230 0.099 0.123 0.197 0.143 [0.126, 0.16] <0.001 24% 0%
40 OpenAI GPT-4-Turbo-2024-04-09 (scratchpad) 0.191 0.206 0.071 0.095 0.198 0.143 [0.126, 0.16] <0.001 23% 0%
41 ForecastBench Imputed Forecaster 0.250 0.107 0.021 0.037 0.179 0.143 [0.129, 0.158] <0.001 34% 100%
42 Anthropic Claude-2.1 (zero shot with freeze values) 0.220 0.184 0.041 0.067 0.202 0.144 [0.124, 0.163] <0.001 31% 0%
43 Google Gemini-1.5-Pro (scratchpad) 0.179 0.233 0.081 0.109 0.206 0.144 [0.125, 0.163] <0.001 24% 0%
44 OpenAI GPT-4 (zero shot) 0.169 0.181 0.105 0.119 0.175 0.144 [0.127, 0.162] <0.001 23% 0%
45 Google Gemini-1.5-Pro (scratchpad with news) 0.181 0.174 0.098 0.112 0.177 0.146 [0.127, 0.166] <0.001 26% 0%
46 OpenAI GPT-4-Turbo-2024-04-09 (scratchpad with news with freeze values) 0.202 0.157 0.076 0.090 0.180 0.146 [0.124, 0.169] <0.001 31% 0%
47 Meta Llama-3-70b-Chat-Hf (zero shot) 0.175 0.233 0.095 0.120 0.204 0.148 [0.128, 0.168] <0.001 24% 0%
48 Anthropic Claude-3-Opus-20240229 (zero shot) 0.185 0.244 0.083 0.112 0.215 0.149 [0.125, 0.173] <0.001 25% 0%
49 Qwen Qwen1.5-110B-Chat (scratchpad with news with freeze values) 0.196 0.182 0.085 0.103 0.189 0.149 [0.132, 0.167] <0.001 25% 0%
50 Mistral AI Mixtral-8x7B-Instruct-V0.1 (zero shot with freeze values) 0.209 0.202 0.065 0.090 0.206 0.150 [0.12, 0.179] <0.001 36% 0%
51 OpenAI GPT-4o (zero shot) 0.205 0.232 0.067 0.097 0.218 0.151 [0.13, 0.171] <0.001 24% 0%
52 Meta Llama-3-8b-Chat-Hf (zero shot with freeze values) 0.201 0.246 0.069 0.101 0.223 0.151 [0.127, 0.175] <0.001 27% 0%
53 OpenAI GPT-4o (superforecaster with news 3) 0.208 0.174 0.076 0.094 0.191 0.151 [0.133, 0.169] <0.001 23% 4%
54 Google Gemini-1.5-Flash (scratchpad) 0.195 0.202 0.086 0.108 0.199 0.151 [0.132, 0.171] <0.001 21% 0%
55 Anthropic Claude-3-Opus-20240229 (superforecaster with news 1) 0.171 0.272 0.101 0.132 0.221 0.151 [0.13, 0.173] <0.001 24% 0%
56 Mistral AI Mistral-Large-Latest (zero shot) 0.173 0.195 0.118 0.132 0.184 0.152 [0.129, 0.176] <0.001 22% 0%
57 Mistral AI Mixtral-8x22B-Instruct-V0.1 (scratchpad with news with freeze values) 0.201 0.181 0.088 0.105 0.191 0.153 [0.133, 0.172] <0.001 23% 0%
58 Qwen Qwen1.5-110B-Chat (scratchpad with freeze values) 0.208 0.205 0.074 0.098 0.206 0.153 [0.136, 0.17] <0.001 21% 0%
59 Mistral AI Mixtral-8x22B-Instruct-V0.1 (scratchpad) 0.190 0.229 0.090 0.116 0.209 0.153 [0.135, 0.17] <0.001 23% 0%
60 Google Gemini-1.5-Pro (superforecaster with news 3) 0.194 0.171 0.101 0.114 0.183 0.154 [0.135, 0.173] <0.001 25% 0%
61 Anthropic Claude-2.1 (scratchpad with news) 0.212 0.190 0.075 0.096 0.201 0.154 [0.135, 0.173] <0.001 26% 8%
62 Google Gemini-1.5-Pro (zero shot) 0.190 0.297 0.079 0.119 0.244 0.154 [0.129, 0.18] <0.001 23% 0%
63 Mistral AI Mixtral-8x7B-Instruct-V0.1 (scratchpad) 0.200 0.131 0.104 0.109 0.166 0.155 [0.136, 0.174] <0.001 28% 14%
64 Qwen Qwen1.5-110B-Chat (scratchpad with news) 0.196 0.188 0.099 0.115 0.192 0.155 [0.137, 0.174] <0.001 24% 0%
65 ForecastBench LLM Crowd (gpt-4o, claude-3.5-sonnet, gemini-1.5-pro) with news 0.224 0.206 0.062 0.088 0.215 0.156 [0.139, 0.172] <0.001 21% 28%
66 Anthropic Claude-3-5-Sonnet-20240620 (superforecaster with news 2) 0.198 0.220 0.090 0.114 0.209 0.156 [0.134, 0.178] <0.001 27% 0%
67 Google Gemini-1.5-Flash (scratchpad with news with freeze values) 0.203 0.232 0.080 0.108 0.218 0.156 [0.133, 0.179] <0.001 24% 0%
68 ForecastBench LLM Crowd (gpt-4o, claude-3.5-sonnet, gemini-1.5-pro) with news 0.225 0.197 0.063 0.087 0.211 0.156 [0.139, 0.173] <0.001 22% 28%
69 Qwen Qwen1.5-110B-Chat (zero shot) 0.196 0.226 0.092 0.116 0.211 0.156 [0.137, 0.175] <0.001 19% 1%
70 ForecastBench LLM Crowd (gpt-4o, claude-3.5-sonnet, gemini-1.5-pro) with news 0.224 0.203 0.063 0.089 0.214 0.156 [0.14, 0.173] <0.001 22% 28%
71 OpenAI GPT-4-Turbo-2024-04-09 (scratchpad with news) 0.202 0.203 0.090 0.110 0.203 0.156 [0.135, 0.178] <0.001 27% 0%
72 OpenAI GPT-4-Turbo-2024-04-09 (superforecaster with news 3) 0.212 0.161 0.087 0.101 0.186 0.157 [0.137, 0.176] <0.001 26% 8%
73 Mistral AI Mixtral-8x22B-Instruct-V0.1 (scratchpad with news) 0.201 0.209 0.092 0.114 0.205 0.157 [0.139, 0.176] <0.001 22% 0%
74 Anthropic Claude-2.1 (scratchpad with news with freeze values) 0.212 0.142 0.094 0.103 0.177 0.157 [0.139, 0.176] <0.001 24% 3%
75 OpenAI GPT-4o (superforecaster with news 1) 0.201 0.255 0.086 0.116 0.228 0.159 [0.133, 0.184] <0.001 27% 0%
76 OpenAI GPT-4-Turbo-2024-04-09 (superforecaster with news 1) 0.197 0.272 0.087 0.121 0.234 0.159 [0.138, 0.181] <0.001 22% 0%
77 Qwen Qwen1.5-110B-Chat (superforecaster with news 1) 0.201 0.303 0.077 0.118 0.252 0.159 [0.136, 0.183] <0.001 23% 0%
78 Anthropic Claude-3-Opus-20240229 (superforecaster with news 3) 0.192 0.129 0.126 0.127 0.161 0.159 [0.139, 0.18] <0.001 22% 5%
79 Google Gemini-1.5-Flash (zero shot) 0.183 0.228 0.115 0.136 0.206 0.160 [0.135, 0.184] <0.001 23% 0%
80 Meta Llama-3-8b-Chat-Hf (scratchpad with freeze values) 0.224 0.170 0.078 0.095 0.197 0.160 [0.145, 0.174] <0.001 23% 0%
81 Meta Llama-3-8b-Chat-Hf (zero shot) 0.201 0.329 0.075 0.121 0.265 0.161 [0.135, 0.187] <0.001 25% 0%
82 Qwen Qwen1.5-110B-Chat (scratchpad) 0.208 0.239 0.087 0.115 0.224 0.161 [0.143, 0.179] <0.001 21% 0%
83 Mistral AI Mistral-Large-Latest (scratchpad with news with freeze values) 0.207 0.200 0.097 0.116 0.203 0.162 [0.141, 0.183] <0.001 24% 0%
84 Meta Llama-3-70b-Chat-Hf (scratchpad) 0.188 0.224 0.116 0.136 0.206 0.162 [0.145, 0.178] <0.001 22% 0%
85 Google Gemini-1.5-Pro (superforecaster with news 1) 0.206 0.267 0.086 0.119 0.236 0.162 [0.141, 0.184] <0.001 26% 0%
86 Anthropic Claude-3-Opus-20240229 (scratchpad with news with freeze values) 0.205 0.212 0.103 0.123 0.208 0.164 [0.142, 0.185] <0.001 24% 0%
87 Anthropic Claude-3-Opus-20240229 (scratchpad with news) 0.205 0.201 0.106 0.123 0.203 0.164 [0.144, 0.185] <0.001 25% 0%
88 Google Gemini-1.5-Flash (scratchpad with news) 0.203 0.203 0.108 0.125 0.203 0.164 [0.142, 0.187] <0.001 24% 0%
89 Anthropic Claude-3-Opus-20240229 (superforecaster with news 2) 0.187 0.243 0.119 0.141 0.215 0.164 [0.141, 0.188] <0.001 24% 0%
90 Mistral AI Mixtral-8x22B-Instruct-V0.1 (superforecaster with news 1) 0.206 0.228 0.103 0.126 0.217 0.166 [0.142, 0.19] <0.001 22% 0%
91 Mistral AI Mixtral-8x22B-Instruct-V0.1 (zero shot) 0.195 0.284 0.105 0.137 0.240 0.166 [0.14, 0.192] <0.001 22% 0%
92 Mistral AI Mistral-Large-Latest (scratchpad with news) 0.207 0.164 0.117 0.126 0.186 0.167 [0.147, 0.186] <0.001 23% 0%
93 Mistral AI Mixtral-8x22B-Instruct-V0.1 (superforecaster with news 3) 0.221 0.168 0.101 0.113 0.194 0.167 [0.15, 0.184] <0.001 19% 12%
94 Mistral AI Mixtral-8x7B-Instruct-V0.1 (zero shot) 0.209 0.233 0.105 0.128 0.221 0.168 [0.14, 0.197] <0.001 25% 0%
95 OpenAI GPT-4o (superforecaster with news 2) 0.232 0.202 0.086 0.107 0.217 0.170 [0.145, 0.194] <0.001 26% 0%
96 Mistral AI Mixtral-8x7B-Instruct-V0.1 (superforecaster with news 1) 0.238 0.207 0.080 0.103 0.222 0.171 [0.148, 0.193] <0.001 29% 15%
97 Mistral AI Mistral-Large-Latest (superforecaster with news 1) 0.205 0.297 0.106 0.141 0.251 0.173 [0.147, 0.198] <0.001 24% 0%
98 Qwen Qwen1.5-110B-Chat (superforecaster with news 3) 0.226 0.207 0.101 0.121 0.216 0.173 [0.156, 0.191] <0.001 23% 4%
99 Anthropic Claude-2.1 (zero shot) 0.220 0.227 0.104 0.127 0.223 0.173 [0.154, 0.192] <0.001 19% 0%
100 Mistral AI Mistral-Large-Latest (superforecaster with news 2) 0.198 0.235 0.130 0.149 0.216 0.174 [0.149, 0.198] <0.001 24% 0%
101 Mistral AI Mixtral-8x7B-Instruct-V0.1 (scratchpad with freeze values) 0.200 0.171 0.147 0.151 0.186 0.176 [0.15, 0.202] <0.001 28% 11%
102 Anthropic Claude-2.1 (superforecaster with news 3) 0.220 0.146 0.132 0.135 0.183 0.177 [0.156, 0.198] <0.001 25% 3%
103 Mistral AI Mixtral-8x22B-Instruct-V0.1 (superforecaster with news 2) 0.232 0.188 0.110 0.124 0.210 0.178 [0.159, 0.198] <0.001 22% 1%
104 Meta Llama-3-8b-Chat-Hf (scratchpad) 0.224 0.229 0.114 0.135 0.227 0.180 [0.163, 0.197] <0.001 22% 0%
105 Google Gemini-1.5-Flash (superforecaster with news 2) 0.210 0.247 0.128 0.150 0.229 0.180 [0.155, 0.205] <0.001 22% 0%
106 Mistral AI Mixtral-8x7B-Instruct-V0.1 (superforecaster with news 2) 0.249 0.185 0.098 0.114 0.217 0.181 [0.155, 0.208] <0.001 32% 22%
107 Mistral AI Mistral-Large-Latest (superforecaster with news 3) 0.233 0.191 0.116 0.129 0.212 0.181 [0.162, 0.201] <0.001 23% 5%
108 Mistral AI Mixtral-8x7B-Instruct-V0.1 (superforecaster with news 3) 0.242 0.166 0.113 0.123 0.204 0.182 [0.162, 0.202] <0.001 28% 13%
109 Meta Llama-2-70b-Chat-Hf (scratchpad with freeze values) 0.224 0.212 0.125 0.141 0.218 0.182 [0.166, 0.199] <0.001 23% 0%
110 OpenAI GPT-4-Turbo-2024-04-09 (superforecaster with news 2) 0.230 0.186 0.126 0.137 0.208 0.183 [0.159, 0.208] <0.001 27% 1%
111 Qwen Qwen1.5-110B-Chat (superforecaster with news 2) 0.223 0.266 0.117 0.144 0.244 0.183 [0.163, 0.204] <0.001 24% 2%
112 Google Gemini-1.5-Flash (superforecaster with news 1) 0.221 0.298 0.116 0.149 0.259 0.185 [0.157, 0.213] <0.001 24% 0%
113 Meta Llama-2-70b-Chat-Hf (zero shot with freeze values) 0.237 0.160 0.129 0.135 0.199 0.186 [0.161, 0.211] <0.001 24% 0%
114 Mistral AI Mixtral-8x7B-Instruct-V0.1 (scratchpad with news with freeze values) 0.278 0.155 0.083 0.096 0.216 0.187 [0.165, 0.208] <0.001 27% 13%
115 Google Gemini-1.5-Flash (superforecaster with news 3) 0.236 0.196 0.126 0.139 0.216 0.187 [0.165, 0.209] <0.001 20% 7%
116 Anthropic Claude-2.1 (superforecaster with news 2) 0.235 0.201 0.129 0.142 0.218 0.188 [0.165, 0.212] <0.001 28% 11%
117 Anthropic Claude-3-Haiku-20240307 (zero shot with freeze values) 0.288 0.152 0.076 0.090 0.220 0.189 [0.171, 0.207] <0.001 26% 0%
118 Anthropic Claude-3-Haiku-20240307 (scratchpad with freeze values) 0.242 0.174 0.131 0.139 0.208 0.191 [0.173, 0.208] <0.001 20% 0%
119 Google Gemini-1.5-Pro (superforecaster with news 2) 0.236 0.280 0.119 0.148 0.258 0.192 [0.163, 0.221] <0.001 22% 0%
120 Meta Llama-2-70b-Chat-Hf (scratchpad) 0.224 0.248 0.149 0.167 0.236 0.196 [0.177, 0.215] <0.001 21% 0%
121 Anthropic Claude-3-Haiku-20240307 (superforecaster with news 2) 0.242 0.187 0.142 0.150 0.214 0.196 [0.177, 0.215] <0.001 21% 0%
122 Anthropic Claude-3-Haiku-20240307 (scratchpad) 0.242 0.214 0.135 0.150 0.228 0.196 [0.178, 0.214] <0.001 22% 0%
123 Mistral AI Mixtral-8x7B-Instruct-V0.1 (scratchpad with news) 0.278 0.197 0.097 0.115 0.238 0.196 [0.173, 0.22] <0.001 26% 13%
124 OpenAI GPT-3.5-Turbo-0125 (scratchpad with freeze values) 0.246 0.274 0.124 0.151 0.260 0.199 [0.18, 0.218] <0.001 22% 0%
125 Anthropic Claude-3-Haiku-20240307 (scratchpad with news with freeze values) 0.265 0.173 0.127 0.135 0.219 0.200 [0.183, 0.217] <0.001 22% 0%
126 Anthropic Claude-3-Haiku-20240307 (scratchpad with news) 0.265 0.202 0.127 0.141 0.233 0.203 [0.185, 0.221] <0.001 22% 0%
127 OpenAI GPT-3.5-Turbo-0125 (scratchpad) 0.246 0.274 0.140 0.165 0.260 0.206 [0.186, 0.225] <0.001 21% 0%
128 Anthropic Claude-3-Haiku-20240307 (zero shot) 0.288 0.174 0.116 0.127 0.231 0.207 [0.188, 0.227] <0.001 23% 0%
129 ForecastBench Always 0.5 0.250 0.250 0.146 0.165 0.250 0.208 [0.198, 0.218] <0.001 17% 0%
130 Anthropic Claude-2.1 (superforecaster with news 1) 0.267 0.258 0.124 0.148 0.263 0.208 [0.184, 0.231] <0.001 24% 3%
131 Meta Llama-2-70b-Chat-Hf (zero shot) 0.237 0.221 0.195 0.200 0.229 0.219 [0.192, 0.245] <0.001 22% 1%
132 Anthropic Claude-3-Haiku-20240307 (superforecaster with news 3) 0.268 0.224 0.168 0.178 0.246 0.223 [0.204, 0.242] <0.001 20% 17%
133 Anthropic Claude-3-Haiku-20240307 (superforecaster with news 1) 0.276 0.243 0.182 0.193 0.260 0.235 [0.209, 0.261] <0.001 21% 0%
134 OpenAI GPT-3.5-Turbo-0125 (zero shot with freeze values) 0.399 0.184 0.067 0.088 0.292 0.244 [0.215, 0.273] <0.001 37% 0%
135 OpenAI GPT-3.5-Turbo-0125 (zero shot) 0.399 0.218 0.163 0.173 0.308 0.286 [0.256, 0.317] <0.001 24% 0%
136 ForecastBench Random Uniform 0.356 0.318 0.207 0.227 0.337 0.292 [0.255, 0.328] <0.001 24% 0%
137 ForecastBench Always 0 0.352 0.500 0.187 0.244 0.426 0.298 [0.245, 0.351] <0.001 29% 0%
138 ForecastBench Always 1 0.648 0.500 0.606 0.587 0.574 0.617 [0.561, 0.673] <0.001 24% 0%