• Korean Journal of Cognitive Science
• /
• v.35 no.1
• /
• pp.23-48
• /
• 2024

Building human-aligned artificial intelligence (AI) for social support remains challenging despite the advancement of Large Language Models. We present a novel method, the Chain of Empathy (CoE) prompting, that utilizes insights from psychotherapy to induce LLMs to reason about human emotional states. This method is inspired by various psychotherapy approaches-Cognitive-Behavioral Therapy (CBT), Dialectical Behavior Therapy (DBT), Person-Centered Therapy (PCT), and Reality Therapy (RT)-each leading to different patterns of interpreting clients' mental states. LLMs without CoE reasoning generated predominantly exploratory responses. However, when LLMs used CoE reasoning, we found a more comprehensive range of empathic responses aligned with each psychotherapy model's different reasoning patterns. For empathic expression classification, the CBT-based CoE resulted in the most balanced classification of empathic expression labels and the text generation of empathic responses. However, regarding emotion reasoning, other approaches like DBT and PCT showed higher performance in emotion reaction classification. We further conducted qualitative analysis and alignment scoring of each prompt-generated output. The findings underscore the importance of understanding the emotional context and how it affects human-AI communication. Our research contributes to understanding how psychotherapy models can be incorporated into LLMs, facilitating the development of context-aware, safe, and empathically responsive AI.

• Proceedings of the Korean Radioactive Waste Society Conference
• /
• 2005.06a
• /
• pp.523-530
• /
• 2005

In this study, maximum exposure rate at DAW(Dry Active Waste) drum surface which is satisfying regulation limit was calculated for conceptual design of IP(Industrial Package). DAW can be classified as combustible and non-combustible waste and the calculation was conducted for single and mixed radionuclide for each type of waste. In case of combustible waste that mixed radionuclide is uniformly distributed, the maximum exposure rates at drum surface were 3.60E-01, 8.85E-01 and 1.27E+01 mSv/hr for IP Type 1, 2-a and 2-b, respectively. and 3.60E-01, 8.85E-01, 1.27E+01 mSv/hr for single radionuclide(Co-60). In case of non-combustible waste that mixed radionuclide is uniformly distributed, the maximum exposure rates at drum surface were 7.14E-01, 1.83E+00, 2.69E+01 mSv/hr for IP Type 1, 2-a and 2-b, respectively. and 7.13E-01, 1.81E-01, 2.62E+01 mSv/hr for single radionuclide(Co-60). Through this study, the maximum amount of DAW can be transported by IP was suggested as maximum exposure rate at drum surface and the calculation for the other types of waste will be conducted.

• Journal of the Korean Chemical Society
• /
• v.31 no.6
• /
• pp.542-554
• /
• 1987

Tetradentate schiff base cobalt(II) complexes; Co(SED), Co(SND) and Co(SOPD) have been prepared, these complexes have react with dry oxygen in DMSO to form oxygen adducts cobalt(III) complexes; $[Co(SED)(DMSO)]_2O_2,\;[Co(SND)(DMSO)]_2O_2$ and $[Co(SOPD)(DMSO)]_2O_2$. It seems to be that the oxygen adducts cobalt(Ⅲ) complexes have heexa coordinated octahedral configration with tetradentate schiff base cobalt (III), DMSO and oxygen, and the mole ratio of oxygen to cobalt(II) complexes are 1 : 2, these complexes have been identified by IR-Spectra, T.G.A., magnetic susceptibilitis and elemental analysis of C.H.N. and Cobalt. The redox reaction process of Co(SED), Co(SND) and Co(SOPD) complexes was investigated by cyclic voltammetry with glassy carbon electrode in 0.1M TEAP-DMSO. The results of redox reaction process of Co(II) / Co(III) and Co(II) / Co(I) for cobalt(SED) and cobalt(SOPD) complexes and Co(II) / Co(III) process for cobalt(SND) complex are reversible process but Co(II) / Co(I) process of Cobalt(SND) complex is irreversible, and oxygen adduct complexes to quasi reversibly with oxygen should be very closed related to the redox potentials of range, $E_{pc}$ = -0.80~-0.89V and $E_{pa}$ = -0.70~-0.76V.

• Korean Journal of Soil Science and Fertilizer
• /
• v.43 no.5
• /
• pp.716-721
• /
• 2010

LCA (Life Cycle Assessment) carried out to estimate carbon footprint and to establish of LCI (Life Cycle Inventory) database of rice production system. The results of collecting data for establishing LCI D/B showed that organic fertilizer and chemical fertilizer input to 4.29E-01 kg $kg^{-1}$ rice and 2.30E-01 kg $kg^{-1}$ rice for rice cultivation. It was the highest value among input for rice cultivation. And direct field emission was 3.23E-02 kg $kg^{-1}$ during rice cropping. The results of LCI analysis focussed on greenhouse gas (GHG) was showed that carbon footprint was 8.70E-01 kg $CO_2$-eq. $kg^{-1}$ rice. Especially for 80% of $CO_2$ in the GHG and 7.02E-01 kg of its $CO_2$-eq. $kg^{-1}$ rice. Of the GHG emission $CH_4$, and $N_2O$ were estimated to be 13% and 5%, respectively. With LCIA (Life Cycle Impact Assessment) for rice cultivation system, it was observed that fertilizer process might be contributed to approximately 80% of GWP (global warming potential).

• Proceedings of the KIEE Conference
• /
• 1997.07f
• /
• pp.2159-2161
• /
• 1997

This paper presents the investigations and the reduction methods for EMI in the Elevator(E/L) system. In the large capacity E/L. Power factor controlled IGBT converter is used, so that the conducted EMI level is mostly determined by a high frequency input current. We suggest the low cost EMI countermearsure for E/L system using by a LCL filter ed a low cost noise filter.

• Journal of the Korean Institute of Electrical and Electronic Material Engineers
• /
• v.25 no.12
• /
• pp.961-968
• /
• 2012

In this study, we have investigated the effects of Mn and Co co-doping on defects, J-E curves and grain boundary characteristics of ZnO-$Bi_2O_3$ (ZB) varistor. Admittance spectra and dielectric functions show two bulk defects of $Zn_i^{{\cdot}{\cdot}}$ (0.17~0.18 eV) and $V_o^{\cdot}$ (0.30~0.33 eV). From J-E characteristics the nonlinear coefficient (${\alpha}$) and resistivity (${\rho}_{gb}$) of pre-breakdown region decreased as 30 to 24 and 5.1 to 0.08 $G{\Omega}cm$ with sintering temperature, respectively. The double Schottky barrier of grain boundaries in ZB(MCo) ($ZnO-Bi_2O_3-Mn_3O_4-Co_3O_4$) could be electrochemically single type. However, its thermal stability was slightly disturbed by ambient oxygen because the apparent activation energy of grain boundaries was changed from 0.64 eV at lower temperature to 1.06 eV at higher temperature. It was revealed that a co-doping of Mn and Co in ZB reduced the heterogeneity of the barrier in grain boundaries and stabilized the barrier against an ambient temperature (${\alpha}$-factor= 0.136).

• Microbiology and Biotechnology Letters
• /
• v.14 no.2
• /
• pp.181-186
• /
• 1986

A defective lambda transducing phase carrying acetyl CoA carboxylase gene (fabE) from Escherichia coli chromosome (72 min on the current linkage map) has been isolated. A restriction map of the chromosomal region from defective transducing phage was established by digestion with combination of the restriction enzymes. No cleavage site for the enzyme EcoRI was found in this region. Restriction fragments were cloned from defective transducing phage into high copy number plasmid vector pACYC184 to generate hybrid plasmids which were capable of complementation of fabE temperature sensitive mutation. We show here that the fabE gene is located on a 3.4 megadalton Bam HI-SalI fragment with a HindIII site, which lies within the 7.4 megadalton BglIIfragment, by complementation analysis.

• Proceedings of the Korean Geotechical Society Conference
• /
• 1997.05a
• /
• pp.147-188
• /
• 1997

• Proceedings of the Korean Society for Rock Mechanics Conference
• /
• 2008.10a
• /
• pp.53-61
• /
• 2008

• Transactions of the Korean Society of Automotive Engineers
• /
• v.18 no.1
• /
• pp.74-85
• /
• 2010

In order to formulate the compounded polypropylene(C-PP) which is suitable to an automotive door trim panel, 9 sorts of properties were measured after manufacturing the C-PP using an extruder and an injection machine with polypropylene(PP), ethylene-octene rubber(EOR) and talc. Mixture design, especially extreme vertices design, in DOE with MINITAB - commercial software was used to analyze the data. The relations between each property and each component, for example, $y=0.00907222x_1+0.00870556x_2+0.0155722x_3$ for specific gravity, were found out by the regression analysis and the variance analysis. The optimized formulation of the C-PP for an automotive door trim panel was acquired at PP(77.6962), EOR(11.0238) and talc(10.2800) by use of the response optimizer(mixture) in MINITAB.